Tacit Agreement Under Section 1 of the Sherman Act

William H. Page

University of Florida Levin College of Law Legal Studies Research Paper Series Paper No. 16-45

1

Tacit Agreement Under Section 1 of the Sherman Act

William H. Page*

Abstract

In dozens of cases each year alleging horizontal price fixing and other per se violations of

Section 1 of the Sherman Act, the central issue is whether the defendants ever formed an

agreement. One source of uncertainty in resolving this issue in litigation is the meaning of “tacit

agreement,” a term the Supreme Court has continued to include within the reach of Section 1, even

as it has emphatically excluded “mere interdependence” or tacit collusion. In this article, I try to

clarify the meaning of tacit agreement and show its practical significance in litigation. After

examining how the Court used the term in Bell Atlantic Corp. v. Twombly, I situate tacit agreement

in the hierarchy of means of coordination, distinguishing it especially from mere interdependence

on the one hand and express agreement on the other. Then I argue for a definition of tacit

agreement—interdependent conduct coordinated by prior private communications of competitive

intentions—and consider what forms of communication and conduct fit that definition. I argue

that, so defined, tacit agreement is more effective than simple interdependence as a means of

coordinating noncompetitive equilibria, and is easier for courts to penalize or enjoin without doing

more harm than good. To show the analytical significance of the concept, I distinguish among four

categories of communications based upon whether the communications are public or private on

the one hand, and whether they relate to present or future conduct on the other. I then examine

cases involving all four kinds of communication to show their relative importance in the

identification and inference of tacit agreement. In the process, I consider the proper meaning and

significance of “signaling” as communication that might form or implement an agreement. The

clarified definition, illustrative cases, and categories of relevant communications will, I argue, help

courts resolve, at every stage of litigation, whether rivals restricted competition by agreement.

* Marshall M. Criser Eminent Scholar, University of Florida Levin College of Law. I thank Joe

Harrington, Danny Sokol, Wentong Zheng and participants at a faculty workshop at the University of

Florida Levin College of Law for comments on various earlier versions of this paper.

2

In perhaps the most frequent confrontation in modern antitrust litigation, plaintiffs allege

that oligopolists have agreed to fix prices,1 allocate markets,2 or exclude rivals,3 all per se violations

of Section 1 of the Sherman Act.4 The defendants respond that, even if they are coordinating their

competitive actions, they have not formed an agreement within the meaning of Section 1;5 they

are only engaging in lawful oligopolistic behavior. When the service stations on Martha’s

Vineyard were accused of fixing prices of gasoline, for example, they pointed out that, because of

their isolated location, relatively small number, and transparent pricing, they could “engage in

‘cooperative pricing’ without any secret meetings or any explicit agreements that would violate

1 In 2015 alone, the Courts of Appeals decided In re Dairy Farmers of America, Inc. Cheese

Antitrust Litigation, 801 F.3d 758, 762–65 (7th Cir. 2015); In re Chocolate Confectionary Antitrust

Litigation, 801 F.3d 383, 395–412 (3d Cir. 2015); and In re Text Messaging Antitrust Litigation, 782 F.3d

867, 872–79 (7th Cir. 2015) (Posner, J.), all concluding the plaintiffs produced insufficient evidence of an

agreement on prices to avoid summary judgment. Three circuit court cases considered the sufficiency of

allegations of agreement on motions to dismiss, with varying outcomes. In re Musical Instruments & Equip.

Antitrust Litig., 798 F.3d 1186, 1193–98 (9th Cir. 2015) (affirming district court’s dismissal); Osborn v.

Visa Inc., 797 F.3d 1057, 1066–69 (D.C. Cir. 2015) (reversing district court’s dismissal); Name.Space, Inc.

v. Interstate Corp. for Assigned Names & Nos., 795 F.3d 1124, 1129–31 (9th Cir. 2015) (affirming district

court’s dismissal). Another, United States v. Apple, Inc., 791 F.3d 290, 321–25 (2d Cir. 2015), affirmed a

trial court’s decision that that Apple had conspired with book publishers to facilitate the adoption of an

agency model of ebook distribution that allowed publishers to raise retail prices. And Ross v. Citigroup,

Inc., 630 F. App’x 79, 82-83 (7th Cir. 2015), affirmed a decision after a bench trial that banks had not

conspired to adopt class-action-barring clauses in credit card agreements.

2 See Stanislaus Food Prods. Co. v. USS-POSCO Indus., 803 F.3d 1084, 1088–95 (9th Cir. 2015)

(finding insufficient evidence that tin can manufacturers had agreed to allocate territories).

3 See Abraham & Veneklasen Joint Venture v. Am. Quarter Horse Ass’n, 776 F.3d 321, 330–34

(5th Cir. 2015) (reversing a jury verdict that defendants conspired to excluded cloned horses from breed

registry); SD3, LLC v. Black & Decker (U.S.) Inc., 801 F.3d 412, 429–35 (5th Cir. 2015) (holding the

complaint sufficiently alleged that table saw manufacturers conspired to exclude plaintiff’s safety

technology); MM Steel, L.P. v. JSW Steel (USA) Inc., 806 F.3d 835, 844–47 (5th Cir. 2015) (reversing a

jury verdict that one steel manufacturer joined a conspiracy of distributors to exclude a rival, but affirming

the verdict against a second manufacturer).

4 15 U.S.C. §1 (2006). On the development and current scope of the per se rule, see 2 JOSEPH

BAUER, WILLIAM H. PAGE & JOHN LOPATKA, FEDERAL ANTITRUST LAW ch. 12 (3d ed. 2013).

5 Section 1 requires a “contract, combination . . . , or conspiracy,” terms that courts have read

collectively to mean agreement. RICHARD A. POSNER, ANTITRUST LAW 262 (2d ed. 2001) (“[T]he courts

sensibly have not worried about whether the terms ‘contract,’ ‘combination,’ and ‘conspiracy,’ in section

1, have nonoverlapping meanings.”).

3

the nation’s antitrust laws. The defendants are each rationally taking account of their competitors’

likely responses to their actions and would be foolish not do so.”6 Courts usually resolve this

confrontation in a casuistic process7 on pretrial motions to dismiss or for summary judgment,

applying the standards of pleading and evidentiary sufficiency to hundreds of patterns of conduct.8

The outcomes on these motions depend in large part on what the courts think a Section 1

agreement is. Even after 125 years of Section 1 litigation, however, the meaning of that

fundamental concept remains uncertain. We do know some things. It is now clear that what the

service stations on Martha’s Vineyard claimed to be doing—“mere interdependence,”9 a process

6 Defendants–Appellees’ Brief, White v. R.M. Packer Co., 635 F.3d 571 (1st Cir. 2011) (No. 10-

1130), 2010 WL 3213231, at *31. The brief then asked rhetorically “[w]ould the plaintiffs prefer that the

defendants not post their prices? Should the defendants not reasonably anticipate the results of their pricing

actions?” Id. at *31–32. The appellate court agreed with the implicit answer. White, 635 F.3d at 585

(“Plaintiffs’ ambiguous evidence is entirely consistent with permissible conscious parallelism.”).

Economists often analyze a similar scenario of gas stations located on opposite corners of the same

intersection in a remote area as an instance of tacit collusion. See LOUIS KAPLOW, COMPETITION POLICY

AND PRICE FIXING 23-24 (2013); ROBERT C. MARSHALL & LESLIE M. MARX, THE ECONOMICS OF

COLLUSION 11–12 (2012); Dennis W. Carlton, Robert H. Gertner & Andrew M. Rosenfield,

Communication Among Competitors: Game Theory and Antitrust, 5 GEO. MASON L. REV. 423, 428–29

(1997).

7 For a discussion of the advantages and disadvantages of casuistic over rule-based decisionmaking,

see Cass R. Sunstein, Problems with Rules, 83 CAL. L. REV. 953, 958 (1995) (“In the casuistic enterprise,

judgments are based not on a preexisting rule, but on comparisons between the case at hand and other cases,

especially those that are unambiguously within a generally accepted norm.”).

8 The decisions on the motions are frequently decisive of the entire litigation. Cases that survive

motions to dismiss on the pleadings may still fail on summary judgment. See, e.g., In re Text Messaging

Antitrust Litig., 782 F.3d 867, 873–79 (7th Cir. 2015) (affirming summary judgment four years after the

same court of appeals had affirmed the denial of a motion to dismiss in In re Text Messaging Antitrust

Litigation, 630 F.3d 622 (7th Cir. 2011)). If a case survives summary judgment, it is frequently settled. The

net effect has been substantially (and controversially) to narrow the range of issues resolved as matters of

fact. See Laumann v. NHL, 56 F. Supp. 3d 280, 307 (S.D.N.Y. 2014) (“It is an unfortunate trend that judges

increasingly resolve trial-worthy disputed fact issues or characterize cases as implausible, thereby disposing

of them on motion rather than allowing them to proceed to trial . . . .”) (citation and internal quotations

omitted).

9 Bell Atl. Corp. v. Twombly, 550 U.S. 544, 557 n.4 (2007) (quoting 6 PHILLIP AREEDA &

HERBERT HOVENKAMP, ANTITRUST LAW ¶ 1425, at 167–85 (2d ed. 2003)).

4

the courts10 and economists11 call tacit collusion12—is not an agreement at all; no matter how much

it restricts output and raises prices, it is per se legal. The Supreme Court explained in 2007 in Bell

Atlantic Corp. v. Twombly13 that the “inadequacy of showing parallel conduct or interdependence,

without more, mirrors the ambiguity of the behavior: consistent with conspiracy, but just as much

in line with a wide swath of rational and competitive business strategy unilaterally prompted by

10 Brooke Grp. Ltd. v. Brown & Williamson Tobacco Corp., 509 U.S. 209, 227 (1993) (“Tacit

collusion, sometimes called oligopolistic price coordination or conscious parallelism, describes the process,

not in itself unlawful, by which firms in a concentrated market might in effect share monopoly power,

setting their prices at a profit-maximizing, supracompetitive level by recognizing their shared economic

interests and their interdependence with respect to price and output decisions.”). See also Motorola Mobility

LLC v. AU Optronics Corp., 775 F.3d 816, 822 (7th Cir. 2014) (dictum) (Posner, J.) (stating if rivals were

to match a slight price increase, it would “be an example of tacit collusion, which is not an antitrust

violation”); White, 635 F.3d at 576 n.3 (“Conscious parallelism has also been called ‘tacit collusion’ or

‘oligopolistic price coordination.’”) (citing Brooke Group, 509 U.S. at 227); In re Insurance Brokerage

Antitrust Litig., 618 F.3d 300, 339 n.19 (3d Cir. 2010) (equating conscious parallelism and tacit collusion

and observing that courts have “found that it is not, without more, sufficient evidence of a § 1 violation,

both because it is not an agreement within the meaning of the Sherman Act, and because it is resistant to

judicial remedies”); Bailey v. Allgas, Inc., 284 F.3d 1237, 1251 (11th Cir. 2002) (“The hallmark of an

oligopoly is tacit collusion among competitors.”).

11 See, e.g., Edward J. Green et al., Tacit Collusion in Oligopoly, in 2 OXFORD HANDBOOK OF

INTERNATIONAL ANTITRUST ECONOMICS 464, 467–68 (Roger D. Blair & D. Daniel Sokol eds., 2015);

Richard A. Posner, Review of Kaplow, Competition Policy and Price Fixing, 79 ANTITRUST L.J. 761, 765

(2014) [hereinafter Kaplow Review] (“What Turner called oligopolistic interdependence . . . I called and

continue to call tacit collusion . . . .”). Posner cited Donald F. Turner, The Definition of Agreement Under

the Sherman Act: Conscious Parallelism and Refusals to Deal, 75 HARV. L. REV. 655 (1962), which did

not actually use the term “oligopolistic interdependence,” but expressed the same idea as “conscious

parallelism.” See, e.g., id. at 671 (“[O]ligopolists who take into account the probable reactions of

competitors in setting their basic prices, without more in the way of ‘agreement’ than is found in ‘conscious

parallelism,’ should not be held unlawful conspirators under the Sherman Act even though . . . they refrain

from competing in price.”).

12 But cf. Patrick Andreoli-Versbach & Jens-Uwe Franck, Econometric Evidence to Target Tacit

Collusion in Oligopolistic Markets, 11 COMPETITION L. & ECON. 463, 467 (2015) (distinguishing “[t]acit

collusion, [which] arises from decisions endogenous to the market by one or several firms that aim to reduce

or eliminate competition” from “oligopolistic interdependence [which] stems from best response to market

conditions, including other firms’ behavior, that favor non-competitive performance”). American antitrust

law makes no such distinction.

13 550 U.S. 544 (2007).

5

common perceptions of the market.”14 Although some scholars, notably Richard Posner15 (until

2014)16 and Louis Kaplow,17 have argued that this sort of coordination should be considered a

Section 1 agreement in some circumstances, the courts have rejected that position. Although courts

recognize that tacit collusion harms consumers, they view it as unavoidable for rational

oligopolists and impossible for courts to remedy without doing more harm than good.18

It is far less clear just what, beyond “interdependence, without more,” constitutes an

agreement.19 Twombly itself said that a sufficient complaint must allege “an agreement, tacit or

express,”20 a phrase the lower courts have quoted dozens of times since 2007. Here, the Court

included within Section 1 not only express agreement, but something called tacit agreement. But

how is including tacit agreement consistent with excluding tacit collusion? Pointing out this

14 Id. at 554.

15 See, e.g., POSNER, supra note 5, at 51–100; Richard A. Posner, Oligopoly and the Antitrust Laws:

A Suggested Approach, 21 STAN. L. REV. 1562, 1563-93 (1968).

16 Kaplow Review, supra note 11, at 763 (“Yet I now think that I didn’t sufficiently appreciate the

force of [Donald] Turner’s doubts about the feasibility of an antitrust remedy for tacit collusion.”) (citing

Turner, supra note 11).

17 Kaplow, supra note 6, at 14 & 16, advocates a “direct approach,” id. at 1, to oligopolistic

interdependence that studies the “nature of the problem, how to detect its presence, and what remedies to

apply,” id. at 13, but declines to advocate a single, optimal rule, which necessarily would depend on

“empirical evidence in realms where existing understanding is incomplete,” id. at 16. See also Reza Dibadj,

Conscious Parallelism Revisited, 47 SAN DIEGO L. REV. 589, 592–93 (2010) (calling for “a more robust,

less anemic conception of antitrust that is willing to confront conscious parallelism”); Thomas A. Piraino,

Jr., Regulating Oligopoly Conduct Under the Antitrust Laws, 89 MINN. L. REV. 9, 23 (2004) (“Tacit

collusion, like express price-fixing, should be illegal on its face under section 1 of the Sherman Act.”).

18 See, e.g., In re Chocolate Confectionary Antitrust Litig., 801 F.3d 383, 397–98 (3d Cir. 2015)

(describing interdependent pricing as a “fact of life in oligopolies” and “courts have no effective remedy

for the problem”) (citations and internal quotations omitted); Clamp-All Corp. v. Cast Iron Soil Pipe Inst.,

851 F.2d 478, 484 (1st Cir. 1988) (interdependent pricing is lawful “not because such pricing is desirable

(it is not), but because it is close to impossible to devise a judicially enforceable remedy for ‘interdependent’

pricing. How does one order a firm to set its prices without regard to the likely reactions of its

competitors?”).

19 KAPLOW, supra note 6, at 149–50 (considering the necessary relationship between definition and

inference).

20 Bell Atl. Corp. v. Twombly, 550 U.S. 544, 553 (2007) (emphasis added) (quoting Theatre Enters.,

Inc. v. Paramount Film Distrib. Corp., 346 U.S. 537, 540 (1954)).

6

puzzle, Judge Cecilia Altonaga wrote recently that the Court’s reference to tacit agreement was

simply a mistake:

It is curious that “tacit agreement” snuck its way into the Twombly decision.

Tacit agreement in an oligopoly is simply conscious parallelism, which the

Supreme Court spent much of the Twombly decision describing as perfectly normal

and not a violation of section 1. Section 1 of the Sherman Act does not reach “tacit

agreement;” there must be an express preceding agreement in order for there to be

a contract, combination, or conspiracy in restraint of trade.21

Others agree with this account of the law.22 For example, the service station operators on Martha’s

Vineyard, in the passage I quoted earlier, contrasted their behavior with “secret meetings or any

explicit agreements that would violate the nation’s antitrust laws.”23 Courts often require that

21 In re Florida Cement & Concrete Antitrust Litig., 746 F. Supp. 2d 1291, 1308 n.13 (S.D. Fla.

2010). Cf. In re Travel Agent Comm’n Antitrust Litig., 583 F.3d 896, 915 (6th Cir. 2009) (Merritt, J.,

dissenting) (criticizing the perceived interpretation of Twombly “to require either an express written

agreement among competitors or a transcribed oral agreement to fix prices”).

22 See, e.g., E.I. DuPont de Nemours & Co. v. FTC, 729 F.2d 128, 143 (2d Cir. 1984) (Lumbard,

J.) (concurring opinion) (lamenting the law’s frequent use of “that accommodating word, ‘tacit,’ which has

created a hole in the agreement requirement large enough at times to swallow it entirely”); Barak Orbach,

The Durability of Formalism in Antitrust, 100 IOWA L. REV. 2197, 2205 (2015) (“The distinction between

concerted action and independent or interdependent conduct is . . . somewhat confusing and counterintuitive

[because] [a]ctual collusions are often tacit, but antitrust law requires plaintiffs to prove that the conspirators

had an explicit agreement.”).

23 Defendants–Appellees’ Brief, White v. R.M. Packer Co., 635 F.3d 571 (1st Cir. 2011) (No. 10-

1130), 2010 WL 3213231, at *31. The court responded that Defendants’ assertion that “‘[a] merely tacit

agreement is not an antitrust violation’ conflates the concepts of ‘tacit collusion,’ referring to bare conscious

parallelism, and ‘tacit agreement,’ which can be reached under § 1, and which plaintiffs allege is in play in

this case.” White, 635 F.3d at 576 n.3 (citation omitted).

7

plaintiffs’ allegations24 or evidence include “plus factors”25 to justify an inference of agreement.26

But most plus factors only “restate the phenomenon of interdependence,”27 by identifying markets

in which oligopolistic coordination is possible, not those in which there is a Section 1 agreement.28

Consequently, the most important plus factor, according to some courts, is noneconomic evidence

of “‘traditional conspiracy,’ or ‘proof that the defendants got together and exchanged assurances

of common action or otherwise adopted a common plan even though no meetings, conversations,

or exchanged documents are shown.’”29

Other scholars, while not ruling out the continued existence of some non-express

agreements have repudiated the term tacit agreement either because of its inherent ambiguity or

24 After Twombly, most courts require allegations of plus factors on motions to dismiss for failure

to state a claim. See, e.g., In re Musical Instruments & Equip. Antitrust Litig., 798 F.3d 1186, 1194 n.7 (9th

Cir. 2015) (Just as circumstantial evidence of plus factors can “raise a genuine issue of material fact to

defeat a defendant's motion for summary judgment, . . . [allegations of plus factors] coupled with parallel

conduct can take a complaint from merely possible to plausible.”). Courts still disagree about how

specifically Twombly requires plaintiffs to plead agreement. See, e.g., Oxbow Carbon & Minerals LLC v.

Union Pac. R.R., 81 F. Supp. 3d 1, 12 (D.D.C. 2015) (acknowledging that while “some courts have

interpreted [Twombly] to require . . . heightened specificity for antitrust conspiracies . . . . [t]his Court . . .

is not persuaded”).

25 For a comprehensive account of plus factors, see William E. Kovacic et al., Plus Factors and

Agreement in Antitrust Law, 110 MICH. L. REV. 393 (2011).

26 See, e.g., In re Flat Glass Antitrust Litig., 385 F.3d 350, 360 (3d Cir. 2004) (observing that “no

exhaustive list [of plus factors] exists,” but identifying three: “(1) evidence that the defendant had a motive

to enter into a price fixing conspiracy; (2) evidence that the defendant acted contrary to its interests; and

(3) ‘evidence implying a traditional conspiracy’”) (citing Petruzzi’s IGA v. Darling–Del. 998 F.2d 1224,

1244 (3d Cir. 1993).

27 Flat Glass, 385 F.3d at 360

28 For example, evidence that defendants (1) had a motive to fix prices or (2) acted against their

individual interest is as consistent with interdependence as it is with agreement. Id. at 360–61. See also In

re Chocolate Confectionary Antitrust Litig., 801 F.3d 383, 398 (3d Cir. 2015) (“[T]hese factors are neither

necessary nor sufficient to preclude summary judgment, at least where the claim is price fixing among

oligopolists.”).

29 Flat Glass, 385 F.3d at 361 (quoting AREEDA & HOVENKAMP, supra note 9, ¶ 1434b, at 243).

See also Chocolate Confectionary, 801 F.3d at 398; In re Titanium Dioxide Antitrust Litig., 959 F. Supp.

2d 799, 823, 829–30 (D. Md. 2013); Kovacic et al., supra note 25, at 396–97 (characterizing as “super plus

factors” those that raise “a strong inference of explicit collusion”).

8

its inconsistent usage in the cases. Gregory Werden notes that “it is counterintuitive that ‘tacit

collusion’ should mean something quite different from ‘tacit agreement.’”30 Louis Kaplow

observes that “[m]any, including the U.S. Supreme Court in both earlier decisions and its most

recent . . . state that tacit agreements are sufficient, yet it is hard to know what to make of these

proclamations given the great ambiguity of the terms.”31 The Areeda and Hovenkamp treatise

“attempt[s] to avoid the ambiguity of ‘tacit agreement’” by not using the term at all.32

Despite these criticisms of substance and terminology, I will show here that tacit

agreement, properly understood, identifies a necessary category of Sherman Act agreement that is

distinguishable from both simple interdependence and express agreement.33 I further suggest that

30 Gregory J. Werden, Economic Evidence on the Existence of Collusion: Reconciling Antitrust

Law With Oligopoly Theory, 71 ANTITRUST L.J. 719, 736 (2004). He adds:

Case law and commentary contrast “tacit” agreements with those that are “express” or

“explicit.” It is doubtful, however, that the distinction between the contrasting terms has

been consistent across cases, or that it was ever the same as that drawn here between

“unspoken” and “spoken” agreements. In the interest of clarity, the descriptors “tacit,”

“express,” and “explicit” all are avoided.”

Id. at 735–36 (citations omitted). Some jurists (such as Judge Altonaga) and commentators still use the

terms synonymously. See, e.g., In re High Fructose Corn Syrup Antitrust Litig., 295 F.3d 651, 654 (7th Cir.

2002) (Posner, J.) (“This statutory language [of Section 1] is broad enough . . . to encompass a purely tacit

agreement to fix prices, that is, an agreement made without any actual communication among the parties to

the agreement.”); Julia Shamir & Noam Shamir, Colluding Under the Radar: Achieving Collusion Through

Vertical Exchange of Information, 63 CLEV. ST. L. REV. 621, 637–38 (2015) (referring both to “illegal tacit

collusion” and “illegal tacit agreement”).

31 KAPLOW, supra note 6, at 45. See also Kovacic et al., supra note 25, at 405 (“The line that

distinguishes tacit agreements (which are subject to section 1 scrutiny) from mere tacit coordination

stemming from oligopolistic interdependence (which eludes section 1’s reach) is indistinct.”).

32 AREEDA & HOVENKAMP, supra note 9, ¶ 1420d, at 143 n.26 (observing that tacit agreement “is

sometimes used for the legal conclusion that an unlawful agreement is present” and “sometimes used to

describe oligopolistic coordination through recognized interdependence alone”; the authors instead “speak

of ‘coordination’ or ‘tacit coordination’” for the economic condition and “‘agreement’ in this context for

the legal conclusion”).

33 See also George A. Hay, Horizontal Agreements: Concepts and Proof, 51 ANTITRUST BULL.

877, 894 (2006). (“[I]f there is to be substance to the concept of a tacit agreement, it must be defined in

such a way that it does not extend to supra-competitive pricing arising purely from oligopolistic

interdependence.”). Hay argues that certain facilitating practices can fill the gap, because “successful

coordination would not be possible (or would be much less likely) without these extra steps—the facilitating

9

a clearer account of tacit agreement can reduce the acknowledged efficiency losses from the

Supreme Court’s immunization of tacit collusion. Courts recognize that simple interdependence is

a form of collusion, but one that oligopolists cannot rationally avoid and that courts cannot

profitably penalize or enjoin. By contrast, tacit agreement, I will show here, requires rivals

consciously to communicate by means that do not benefit consumers and that courts can practically

identify and deter in litigation. Tacit agreement is also more likely to be successful in coordinating

output and prices than simple interdependence.

The term is admittedly ambiguous, but it is the one the courts use. And it is no more

ambiguous than any other shorthand designation of non-express agreements we might propose—

such as “concerted action.”34 Nor is its ambiguity or inconsistency in usage cause for despair:

courts have, by necessity35 and design,36 given content to equally ambiguous antitrust terms

practices.” Id. at 900. But cf. William H. Page, Facilitating Practices and Concerted Action Under Section

1 of the Sherman Act, in ANTITRUST LAW AND ECONOMICS 23, 34 (Keith Hylton ed., 2009) (“[I]ndustry-

wide adoption of [a facilitating] practice alone would likely not justify an inference of a horizontal

agreement, because the pattern would be consistent with lawful oligopolistic interdependence.”).

34 I have elsewhere examined non-express coordination as concerted action. See, e.g., William H.

Page, Objective and Subjective Theories of Concerted Action, 79 ANTITRUST L.J. 215 (2013); William H.

Page, Communication and Concerted Action, 38 LOY. U. CHI. L. J. 405 (2007). This is a term I propose as

an analogy to concerted practices under Article 101 of the EU Treaty. Consolidated Version of the Treaty

on the Functioning of the European Union art. 101, May 9, 2008, 2008 O.J. (C 115) 47. See also Federico

Ghezzi & Mariateresa Maggiolino, Bridging EU Concerted Practices with U.S. Concerted Actions, 10 J.

COMPETITION L. & ECON. 647 (2014) (developing the analogy further). But the term is also ambiguous in

that context, because courts use it to embrace all forms of Sherman Act agreement. See, e.g., Am. Needle,

Inc. v. NFL, 560 U.S. 183, 190 (2010) (“Section 1 applies only to concerted action that restrains trade.”);

In re Chocolate Confectionary Antitrust Litig., 801 F.3d at 395 (3d Cir. 2015) (“Without proof of concerted

action, the plaintiff’s claim fails because the very essence of a section 1 claim . . . is the existence of an

agreement.”) (citations and internal quotations omitted).

35 Larry Kramer, The Lawmaking Power of the Federal Courts, 12 PACE L. REV. 263, 269 (1992)

(“[C]ourts must make a certain amount of common law simply because there is no clear line between

‘making’ and ‘applying’ law, between commands that are clear on the face of a statute and those made

through an exercise of judgment and creativity.”).

36 The drafters of the Sherman Act assumed that ambiguous terms and categories can gain clarity

by a process of reconceptualization and application. Senator Sherman said that the meaning of the vague

categories of the proposed Sherman Act itself “must be left for the courts to determine in each particular

case.” See 21 Cong. Rec. 2460 (1890) (remarks of Sen. Sherman). See also Frank H. Easterbrook, Statutes’

10

throughout the Sherman Act’s history, often with the aid of legal and economic scholarship.37

Think, for example, of “predatory pricing,”38 “essential facilities,”39 or “antitrust injury.”40 The

task for tacit agreement is no different: to frame an economically justified and legally administrable

standard.

The next Part begins by drawing from Twombly four categories of parallel oligopolistic

behavior—independent conduct, interdependent conduct, tacit agreement, and express agreement.

I argue that tacit agreement differs from simple interdependence or tacit collusion in that rivals

coordinate their interdependent conduct by economically inefficient communication of their

intentions on critical competitive choices facing the firms; it differs from express agreement in that

the communications do not involve mutual assurances that amount to a completed agreement that

precedes any interdependent conduct. Although it may seem paradoxical to call coordination that

involves communication a tacit agreement, the idea is that rivals demonstrate their accord with a

Domains, 50 U. CHI. L. REV. 533, 544 (1983) (the Sherman Act “effectively authorize[s] courts to create

new lines of common law”).

37 The Supreme Court has even “felt relatively free to revise our legal analysis as economic

understanding evolves and . . . to reverse antitrust precedents that misperceived a practice’s competitive

consequences.” Kimble v. Marvel Entm’t, LLC, 135 S. Ct. 2401, 2412–13 (2015). Even when precedents

remain nominally in place, changing theory can change their meaning. See generally William H. Page, The

Chicago School and the Evolution of Antitrust: Characterization. Antitrust Injury, and Evidentiary

Sufficiency, 75 VA. L. REV. 1221, 1228 (1989) (“The Court’s use of [Chicago School] models permits it to

reflect prevailing economic understanding, while sharing responsibility for the large-scale estimates of

legislative fact on which substantive rules are based. Moreover, it allows the courts to resolve important

factual issues in individual cases, and permits knowledge about antitrust practices to develop through the

ongoing process of litigation.”).

38 For example, in Brooke Group Ltd. v. Brown & Williamson Tobacco Corp., 509 U.S. 209, 225

(1993), the Supreme Court adopted elements of the Areeda-Turner test for predatory pricing. Phillip Areeda

& Donald F. Turner, Predatory Pricing and Related Practice Under Section 2 of the Sherman Act, 88

HARV. L. REV. 697, 698 (1975).

39 See, e.g., Phillip Areeda, Essential Facilities: An Epithet in Need of Limiting Principles, 58

ANTITRUST L.J. 841(1990).

40 See, e.g., Ronald W. Davis, Standing on Shaky Ground: The Strangely Elusive Doctrine of

Antitrust Injury, 70 ANTITRUST L.J. 697, 750 (2003); William H. Page, The Scope of Liability for Antitrust

Violations, 37 STAN. L. Rev. 1445, 1459–83 (1985).

11

communicated proposal or expression of intention not by a verbal acceptance, but by subsequent

interdependent conduct. There is no “preceding agreement;” the combination of communications

and conduct form the tacit agreement. In Parts III through VI, I will consider what sorts of

communications that coordinate interdependent behavior can form a tacit agreement. I examine

cases that illustrate four categories of communications with varying degrees of probative value in

the inference of tacit agreement, depending upon whether the communications are private or public

and whether they relate to present or future choices. I also explain the relationship of these

categories of communication to “signaling” a term that often appears in the cases I am

considering.41 With a clarified understanding of tacit agreement, courts will be better able to

evaluate patterns of evidence in resolving issues of agreement in every procedural setting.42

I. DEFINING AND INFERRING TACIT AGREEMENT

Twombly recognized tacit agreement as a form of coordination somewhere between

interdependence and express agreement. Unfortunately, it confused matters both by what it said

and did not say about these categories in its discussion of the relationship between the agreement

requirement and the pleading standard. In this Part, I examine the Court’s presentation of its

categories of coordination, then reconstruct and try to justify a more coherent definition of tacit

41 Porter defines a signal as an “action by a competitor that provides a direct or indirect indication

of its intentions, motives, goals, or internal situation.” MICHAEL E. PORTER, COMPETITIVE STRATEGY:

TECHNIQUES FOR ANALYZING INDUSTRIES AND COMPETITORS 75 (1980). See also Oliver Heil & Thomas

S. Robertson, Toward a Theory of Competitive Market Signaling: A Research Agenda, 12 STRATEGIC

MGMT. J. 403, 403 (1991) (“[C]ompetitive reactions are frequently based on signals which precede actual

actions in the marketplace.”).

42 See Hay, supra note 33, at 895. (“Without a working definition of a tacit agreement and an

understanding of how it differs from pure oligopolistic interdependence, there is no basis on which a jury

can decide whether or not a tacit agreement has occurred or why the hypothesis of a tacit agreement is more

plausible than the hypothesis of pure oligopolistic interdependence.”). The same concerns arise in the

evaluation of allegations on a motion to dismiss and the evaluation of evidence on a motion for summary

judgment.

12

agreement. In the last section, I suggest that the relevance of particular communications to tacit

agreement depends, first, on whether the communications relate to present or future competitive

conditions, and second on whether the communications are addressed to multiple audiences or

only to rivals.

A. TWOMBLY’S CONCEPT(S) OF AGREEMENT

In Twombly, the Supreme Court told us that the “‘crucial question’ is whether the

challenged anticompetitive conduct ‘stem[s] from independent decision or from an agreement,

tacit or express.’”43 It added in the same paragraph that “[e]ven ‘conscious parallelism,’ a common

reaction of ‘firms in a concentrated market [that] recogniz[e] their shared economic interests and

their interdependence with respect to price and output decisions is ‘not in itself unlawful.’”44 The

Court quoted scholars who emphasize that “mere interdependent parallelism”45 and “mere

interdependence of basic price decisions”46 are not agreements. In this passage, the Court

recognized a spectrum of four categories of parallel oligopolistic behavior: (1) independent

conduct, in which rivals act in the same way regardless of one another’s actions; (2) “mere”

interdependent conduct, in which rivals act in the same way only because each expects the others

to do so; (3) tacit agreement; and (4) express agreement. The first two categories of conduct are

not agreements under Section 1; the latter two are. Twombly also stressed the need to avoid “false

inferences from identical behavior,” and so required the complaint to allege “enough factual matter

(taken as true) to suggest that an agreement was made.”47 The conduct alleged in Twombly itself—

43 Bell Atl. Corp. v. Twombly, 550 U.S. 544, 553 (2007) (emphasis added) (quoting Theatre Enters.,

Inc. v. Paramount Film Distrib. Corp., 346 U.S. 537, 540 (1954)).

44 Id. at 553–54 (citing Brooke Group, 509 U.S. at 227).

45 Id. at 554 (quoting AREEDA & HOVENKAMP, supra note 9, ¶ 1433a, at 236).

46 Id. (quoting Turner, supra note 11, at 672).

47 Id. at 556.

13

parallel efforts to disadvantage new entrants48 and parallel choices not to enter one another’s

territory49—could easily be explained as independent or simply interdependent choices, so the

complaint was insufficient to raise an inference of either kind of agreement.

In this formulation, the plaintiff must allege facts about the defendants’ conduct that

suggest coordination by something more than mere interdependence. Oddly, however, the Court

did little to explain what more, beyond interdependence, would minimally justify the inference of

a tacit agreement. We can infer from the distinction between mere interdependence and tacit

agreement that the latter category requires rivals to coordinate their actions by distinctive means,

presumably involving a different kind of communication, but the Court offered no indication of

what means of coordination might be enough. It did suggest that allegations of parallel conduct

are insufficient “without that further circumstance pointing toward a meeting of the minds,”50 but

that tells us nothing useful, because rivals can reach a “meeting of the minds” by simple

interdependence. Kaplow points out that the term meeting of the minds “readily covers behavior

that is interdependent . . . , such as the standard scenario in which firms in an oligopoly are able to

48 Id. at 566 (reasoning that parallel actions by the incumbent carriers to hinder rivals’ entry raised

no plausible inference of agreement, because the conduct could well have been “the natural, unilateral

reaction of each ILEC intent on keeping its regional dominance”).

49 Id. at 567–68 (reasoning that, because of the defendants’ history as regulated monopolies, a

“natural explanation” for their failure to enter one another’s territories was that they “were sitting tight,

expecting their neighbors to do the same thing”). Compare the allegations of territorial allocation in

Twombly with Northstar Energy LLC v. Encana Corp., No. 1:13-CV-200, 2014 WL 5343423, at *5 (W.D.

Mich. Mar. 10, 2014) (denying a motion to dismiss where mails between the defendants “attached to the

complaint reveal their discussions on how to avoid ‘bidding each other up’ in many prospective lease deals

and at a public land auction” and “discussed extensively how to divide up the counties”).

50 Twombly, 550 U.S. at 557. Lower courts continue to refer to the meeting of the minds formulation

as if it distinguishes agreement from interdependent action. See, e.g., GSI Tech. Inc. v. Cypress

Semiconductor Corp., No. 11-CV-03613, 2015 WL 365491, at *5 (N.D. Cal. Jan. 27, 2015) (“So long as

the parties, in a meeting of the minds, coordinated horizontal behavior with the purpose of committing a

violation, they remain exposed.”).

14

coordinate their prices by understanding each other’s thought processes, which forms the basis for

predicting their reactions to different prices that each firm may charge.”51

We can also infer from Twombly’s reference to “an agreement, tacit or express” that the

former kind of agreement is different from the latter. The term “express agreement” presumably

means in this usage what it usually means in law—a completed, specified agreement formed by

promises in language or equivalent signifiers.52 As I mentioned in the introduction, courts on

summary judgment motions often look for evidence tending to show “traditional conspiracy,” an

agreement “in which the defendants got together and exchanged assurances of common action.”53

By implication, a tacit agreement is one that does not involve those sorts of assurances. But the

Court confused matters here as well by stating, in a frequently quoted passage, that “when

allegations of parallel conduct are set out in order to make a § 1 claim, they must be placed in a

51 KAPLOW, supra note 6, at 34. See also See Jonathan B. Baker, Two Sherman Act Section 1

Dilemmas: Parallel Pricing, the Oligopoly Problem, and Contemporary Economic Theory, 38 ANTITRUST

BULL. 143, 178 (1993) (“[I]f an agreement is defined as a meeting of the minds, a court conscientiously

applying the definition would likely infer an agreement from the consciously parallel interaction among

oligopolists. When firm 1 raises price, expecting firm 2 to do the same, and firm 2 rises price expecting

firm 1 to likewise, the firms have reached a common understanding by communicating solely through their

pricing actions.”).

52 See, e.g., BLACK’S LAW DICTIONARY 303 & 701 (10th ed. 2014) (defining “express contract” as

“[a] contract whose terms the parties have explicitly set out,” and “express” as “[c]learly and unmistakably

communicated; stated with directness and clarity”); Rosenfeld v. Brooks, 895 S.W.2d 132, 135 (Mo. Ct.

App. 1995) (an “express contract requires the showing of explicit promises made by the parties”).

Twombly itself quoted the district court’s reference to Theatre Enterprises’ statement that

“[c]ircumstantial evidence of consciously parallel behavior may have made heavy inroads into the

traditional judicial attitude toward conspiracy; but ‘conscious parallelism’ has not yet read conspiracy out

of the Sherman Act entirely.” Twombly, 550 U.S. at 552 (quoting Theatre Enters., Inc. v. Paramount Film

Distrib. Corp., 346 U.S. 537, 259–60 (1954)). For this statement, Theatre Enterprises cited a law review

article that described “traditional conspiracy requirements” that “[t]he big cases of the first fifty years of

the Sherman Act readily met,” because they involved formal associations that adopted explicit cartel

measures as rules. James A. Rahl, Conspiracy and the Anti-Trust Laws, 44 ILL. L. REV. 743, 752 (1950).

53 In re Flat Glass Antitrust Litig., 385 F.3d 350, 361 (quoting AREEDA & HOVENKAMP, supra note

9, ¶1434b, at 243). See also Baker, supra note 51, at 179 (observing that agreement is a “process to which

the law objects: a negotiation that concludes when the firms convey mutual assurances that the

understanding they reached will be carried out”).

15

context that raises a suggestion of a preceding agreement.”54 But a “preceding” agreement implies

that the parties have already formed an express, verbal agreement, which they then implemented

by acting interdependently.55 To require a preceding agreement seems to require an exchange of

promises and to exclude the possibility of a tacit agreement formed in part by the conduct itself.

One reading of these perplexing passages is that the Court’s references to tacit agreement

were careless oversights (or perhaps anachronisms)56 and there is no longer such a category of

agreement. In his summary judgment decision in Text Messaging, Judge Posner wrote that

“[e]xpress collusion violates antitrust law; tacit collusion does not,”57 apparently not admitting any

intermediate category of tacit agreement.58 As I noted in the introduction, Judge Altonaga took

tacit agreement to mean mere interdependent conduct. This view assumes that the exclusion of

54 Twombly, 550 U.S. at 557 (emphasis added). See, e.g., In re Musical Instruments & Equip.

Antitrust Litig., 796 F.3d 1186, 1194 (9th Cir. 2015).

55 See, e.g., In re Florida Cement & Concrete Antitrust Litig., 746 F. Supp. 2d 1291, 1308 n.13

(S.D. Fla. 2010) (Altonaga, J.).

56 The Court’s phrase “agreement, tacit or express” was a quotation from Theatre Enterprises, 346

U.S. at 540. That case approvingly cited American Tobacco Co. v. United States, 328 U.S. 781 (1946),

which had inferred an unlawful agreement solely from identical price changes that were unjustified by

demand and cost conditions, id. at 810, conduct modern courts would characterize as mere interdependence.

As Donald Turner later wrote, the “facts in American Tobacco were consistent with [the] hypothesis” of

interdependent oligopoly pricing “without overt communication or agreement, but solely through a rational

calculation by each seller of what the consequences of his price decision would be, taking into account the

probable or virtually certain reactions of his competitors.” See Turner, supra note 11, at 661. Where pricing

is interdependent and “each seller is fully aware of the interdependence and the consequences of his taking

advantage of it,” (as in American Tobacco, but not Theater Enterprises itself) “it is not at all preposterous

. . . to classify what transpires as a ‘tacit agreement.’” Id. at 663. It is conceivable, then, that Theatre

Enterprises eight years later meant to include that sort of conduct in tacit agreement. Regardless, Twombly

has made clear that, now, mere interdependence is lawful oligopoly behavior.

57 In re Text Messaging Antitrust Litig., 782 F.3d 867, 872 (7th Cir. 2015).

58 In an earlier case, he assumed a distinction between a lawful “purely tacit agreement” and an

unlawful “express, manifested agreement.” In re High Fructose Corn Syrup Antitrust Litig., 295 F.3d 651,

654 (7th Cir. 2002). Courts continue to quote Posner’s statement after Twombly. See, e.g., Fleischman v.

Albany Med. Ctr., 728 F. Supp. 2d 130, 157 (N.D.N.Y. 2010); In re Ethylene Propylene Diene Monomer

(EPDM) Antitrust Litig., 681 F. Supp. 2d 141, 166 (D. Conn. 2009).

16

mere interdependence from tacit agreement left the term with no content and that all that remains

in the agreement category are express agreements.

Many other courts, however, have assumed that Twombly intentionally included tacit

agreements within Section 1 and meant to distinguish a category of conduct from both express

agreement and mere interdependence. For example, in White v. R.M. Packer Co.,59 the court

insisted that Twombly “reiterated that tacit agreements are still agreements under the Sherman

Act.”60 The court rejected the defendants’ contrary contention because it “conflated the concepts

of ‘tacit collusion,’ referring to bare conscious parallelism, and ‘tacit agreement,’ which can be

reached under § 1.”61 Many post-Twombly courts still quote a 1987 decision of the Second Circuit

that defined “mere interdependent behavior” as “actions taken by market actors who are aware of

and anticipate similar actions taken by competitors, but which fall short of a tacit agreement.”62

So, again, what is a tacit agreement?

B. TACIT AGREEMENT AND PRIVATE COMMUNICATION

Post-Twombly cases that recognize the category of tacit agreement have usually not tried

to define it. In some cases, they do not need to define it, because the plaintiff alleged or offered

59 635 F.3d 571 (1st Cir. 2011).

60 Id. at 576.

61 Id. at 576 n.3 (citing Twombly’s quotation of Theatre Enterprises). See also Beltran v.

InterExchange, Inc., No. 14-cv-03074-CMA-KMT, 2016 WL 1253622, at *3 n.7 (D. Colo. Mar. 31, 2016)

(quoting White’s “helpful explanation of the difference between a tacit agreement, which is prohibited under

the Sherman Act, and . . . ‘conscious parallelism’”); In re Text Messaging Antitrust Litig., No. 08 C 7082,

2009 WL 5066652, at *5 (N.D. Ill. Dec. 10, 2009) (quoting Posner’s statement that defendants must have

formed “an actual, manifest agreement not to compete,” but noting that Twombly “allows for the possibility

of a ‘tacit’ agreement”).

62 Apex Oil Co. v. DiMauro, 822 F.2d 246, 254 (2d Cir. 1987). The court in Apex evidently thought

that tacit agreement involved some form of communication. See id. at 257 (finding an agreement “in light

of all of the circumstances, including communications among the defendants and their parallel conduct”

and reversing lower court’s grant of summary judgment in favor of a single defendant).

17

evidence of only independent or unilateral conduct63 or because the plaintiff provided sufficient

allegations or evidence of express agreement.64 In some cases, courts hold that the evidence is

sufficient to infer an agreement without ever making clear what sort of agreement they mean.65

Courts have instructed juries, unhelpfully, that meetings and discussions among rivals are neither

necessary nor sufficient to prove a conspiracy; what matters is whether “the alleged members of

the conspiracy in some way came to an agreement to accomplish a common purpose.”66 This

circular formulation never clarifies what communications can form an agreement and, of course,

fails to distinguish express and tacit agreements 67

63 See, e.g., Mayor & City Council of Balt. v. Citigroup, Inc., 709 F.3d 129, 139–40 (2d Cir. 2013)

(citing Apex, but concluding that defendants’ “en masse flight from a collapsing market in which they had

significant downside exposure—made perfect business sense”); Kendall v. Visa U.S.A., Inc., 518 F.3d

1042, 1047–48 (9th Cir. 2008) (citing Twombly’s reference to tacit agreement, but holding that complaint

“failed to plead any evidentiary facts beyond parallel conduct to prove their allegation of a conspiracy”).

64 See, e.g., SD3, LLC v. Black & Decker (U.S.) Inc., 801 F.3d 412, 430 (4th Cir. 2015) (holding

that complaint’s “detailed story” of defendant’s boycott of plaintiff’s technology sufficiently alleged an

express agreement); Arapahoe Surgery Ctr. v. Cigna Helathcare, Inc., 80 F. Supp. 3d 1257, 1264 (D. Colo.

2015) (finding allegations sufficient to withstand a motion to dismiss where plaintiff alleged that “Cigna

representatives joined conspiratorial meetings on [two specified dates], at which the agreement was

confirmed and actions were planned, and that Cigna's participation in the agreement and acts in furtherance

thereof were mentioned in e-mails exchanged between co-conspirators” between those dates).

65 See, e.g., City of Tuscaloosa v. Harcros Chems., Inc., 158 F.3d 548 (11th Cir. 1998). The court

there held that “[o]ligopolists behaving in a legal, consciously parallel fashion could achieve high and rising

prices, even as costs remained stable, by engaging in price leadership [, but the] odds that they could achieve

a price and profit increase and maintain incredibly high incumbency rates—that is, maintain the very same

distribution of municipal contracts year after year—are minuscule . . . unless the oligopolists were

communicating with one another.” Id. at 572. Consequently, “[i]t would be permissible for a jury to find .

. . that the defendants were not in fact competing for contracts but were concertedly protecting each other’s

incumbencies and pushing up prices.” Id. at 573. At no point did the court specify what sorts of

communications or the nature of the agreement that the jury could infer.

66 Toledo Mack Sales & Serv., Inc. v. Mack Trucks, Inc., 386 F. App’x 214, 221 (3d Cir. 2010).

This instruction is drawn from ABA SECTION OF ANTITRUST LAW, MODEL JURY INSTRUCTIONS IN CIVIL

ANTITRUST CASES, 2005 EDITION B-2–B-3 (2005).

67 For related criticism of the model instructions, see Louis Kaplow, On the Meaning of Horizontal

Agreements in Competition Law, 99 CAL. L. REV. 683, 752–54 (2011); William H. Page, Communication

and Concerted Action, 38 LOY. U. CHI. L.J. 405, 420–23 (2007).

18

Other courts seem to think that a non-express agreement is one formed by nonverbal

signals. According to one court, “[t]he law forbids nonverbal agreements in restraint of trade, as

well as express ones; otherwise, the law would be emasculated as competitors accomplished the

forbidden goal with a wink and a nod.”68 The court here appears to use “express” to mean oral or

written; that would seem to imply that one formed by gestures rather than words is not express,

but only tacit. But an agreement formed by the intentional use of a well-understood conventional

signal like a nod can express assent and complete an agreement in the same way as words.69 Courts

have affirmed drug conspiracy convictions on the basis of testimony that a defendant nodded

during a conspiratorial meeting.70 There are apparently no antitrust cases to the same effect,71

perhaps because antitrust conspiracies are more complex than the typical drug deal, and

conventional signals usually cannot convey much more information than assent or dissent. But,

conceptually, any conventional signifier that unmistakably conveys the necessary assurance, like

68 Blomkest Fertilizer, Inc. v. Potash Corp. of Sask., 176 F.3d 1055, 1063 (8th Cir. 1992),

superseded en banc on other grounds, 203 F.3d 1028, 1037 (8th Cir. 2000). Esco recognized that “[a]

knowing wink can mean more than words.” Esco Corp. v. United States, 340 F.2d 1000, 1007 (9th Cir.

1965). Cf. Evans v. United States, 504 U.S. 255, 274 (1992) (Kennedy, J., concurring) (parties to a bribe

“need not state the quid pro quo in express terms, for otherwise the law’s effect could be frustrated by

knowing winks and nods”); Meyer v. Kalanick, 15 Civ. 9796, 2016 WL 1266801, at *5 (S.D.N.Y. Mar. 13,

2016) (“Sophisticated conspirators often reach their agreements as much by the wink and the nod as by

explicit agreement, and the implicit agreement may be far more potent, and sinister, just by virtue of being

implicit.”).

69 See ADAM KENDON, GESTURE: VISIBLE ACTION AS UTTERANCE 2 (2004) (“[D]etailed studies .

. . have shown that” gesture and speech “are so intimately connected that they appear to be governed by a

single process.”); KAPLOW, supra note 6, at 67 (hypothesizing the use of a thumbs-up signal to form an

agreement).

70 United States v. Grajeda-Encinas, 474 F. App’x 506, 508 (9th Cir. 2012) (holding that where

defendant “was observed giving an authoritative nod to a co-conspirator regarding the amount of cocaine

in their possession . . . [, t]he jury was entitled to conclude that he exercised dominion or control over the

cocaine or that he participated in a ‘joint venture’ to possess the drugs”); United States v. Torres, 114 F. 3d

520, 525 (5th Cir. 1997) (holding that evidence that defendant was present at a meeting and “nodded his

head during the discussion of transporting drugs” was “barely” sufficient to support a conviction).

71 One found that some of the realtors at a meeting “gave the impression that his firm would adopt

a similar” commission increase. United States v. Foley, 598 F.2d 1323, 1332 (4th Cir. 1979). The court

may have been suggesting that those realtors gave that impression by a nonverbal signal.

19

a nod or a handshake, can form an express, completed agreement. Similarly, conspirators may

develop a jargon that allows them to communicate express assent by different and fewer words

than they would in ordinary language.72 In all of these cases, the agreements are as express as any.

A tacit agreement is better understood as one in which rivals communicate their intentions

in language without forming a complete agreement, but then indicate their assent to the suggested

course of action by subsequent interdependent pricing or other competitive actions. The First

Circuit in White v. R.M. Packer Co.,73 offered a good starting point for such a definition of tacit

agreement: “‘uniform behavior among competitors, preceded by conversations implying that later

uniformity might prove desirable or accompanied by other conduct that in context suggests that

each competitor failed to make an independent decision.’”74 The passage is admittedly problematic

in a couple of ways. First, it is a quotation from the Supreme Court’s decision Brown v. Pro

Football, Inc.,75 which dealt with antitrust’s labor exemption and never used the term tacit

72 See, e.g., In re High Fructose Corn Syrup Antitrust Litig., 295 F.3d 651, 662 (7th Cir. 2002) (“A

director . . . was reported to have said that ‘every business I’m in is an organization’ (emphasis added)—

which sounds innocuous enough, but he said it in reference to the conspiracy to fix the price of lysine

(‘lysine is an organization’) and so in context it appears that ‘organization’ meant price-fixing

conspiracy.”); United States v. David E. Thompson, Inc., 621 F.2d 1147, 1152 (1st Cir. 1980) (referring to

an “unchanging jargon” in the conspiracy); United States v. Rodgers, 624 F.2d 1303, 1308 (5th Cir. 1980)

(conspirators other than the designated winning bidders were “given prices ‘to clear’ or ‘to protect’ and

either submitted complementary bids or did not bid at all); United States v. Consol. Packaging Corp., 575

F.2d 117, 121 (7th Cir. 1978) (“Conspirators had their own helpful jargon. Those conspirators who were

mutually cooperative during any particular period were referred to as being ‘on the phone,’ and those who

were not conspirators, or were at least not active for a particular period were referred to as ‘off the phone.’”).

73 635 F.3d 571 (1st Cir. 2011).

74 Id. at 576 (quoting Brown v. Pro Football, Inc., 518 U.S. 231, 241 (1996)). A number of courts

have quoted this definition. See In re Nexium (Esomeprazole) Antitrust Litig., 42 F. Supp. 3d 231, 250 (D.

Mass. 2014) (quoting White’s tacit agreement standard); In re Online Travel Co. (OTC) Hotel Booking

Antitrust Litig., 997 F. Supp. 2d 526, 535 & n.10 (N.D. Tex. 2014) (quoting White’s distinction between

interdependence and tacit agreement and quoting its reference to Interstate Circuit as an illustration of the

latter).

75 518 U.S. at 241.

20

agreement.76 Moreover, the second clause in the passage—conduct that suggests the rivals did not

act independently—seems to include simple interdependent conduct. White, in quoting this

passage, obviously did not mean to include mere interdependence in tacit agreement, because it

ultimately rejected plaintiff’s claim that the interdependent conduct of the gas stations on Martha’s

Vineyard was unlawful.

But the first clause in the passage is more helpful: “uniform behavior among competitors,

preceded by conversations implying that later uniformity might prove desirable.” This passage

contemplates a scenario in which rivals agree by acting uniformly in response to earlier

communications that do not themselves form a verbal agreement.77 Implicit in the passage are the

assumptions that, first, the uniform behavior is interdependent and, second, the antecedent

“conversations” consist of communications that convey the rivals’ intent, without an express

exchange of assurances. The Supreme Court in Brown cited United States v. Foley78 as an example

of the sort of agreement described in the quoted passage. In that case, rivals met privately,

discussed the need for a price increase, and each stated or indicated his intent to raise his price.

Although the rivals in Foley never exchanged promises to raise prices, the court affirmed criminal

convictions. I discuss Foley further in Part III, but the crucial point here is its recognition of a

category of non-express agreement, in which rivals clarify their expectations about one another’s

intentions by communication, then act consistently with the communications. In Esco Corp. v.

76 In the passage, the Court was describing, not endorsing, some theories of antitrust liability that

might threaten labor unions. It introduced the passage with the grudging observation that antitrust

“sometimes permits judges or juries to premise antitrust liability upon little more than” the described

conduct, without spelling out what “more” courts actually require. Id. at 241.

77 See, e.g., Jacobs v. Tempur-Pedic Int’l, 626 F.3d 1327, 1343 (11th Cir. 2010) (suggesting the

plaintiff must allege “that, in addition to tacitly colluding, [rivals] signaled each other on how and when to

maintain or adjust prices”).

78 598 F.2d 1323 (4th Cir. 1979).

21

United States,79 the court hypothesized essentially the same scenario80 and held that it could

amount to sufficient evidence of an agreement in a criminal case. The court added that an “express

agreement” with written or oral assurances is not required, “if a course of conduct, or a price

schedule, once suggested or outlined by a competitor in the presence of other competitors, is

followed by all—generally and customarily—and continuously for all practical purposes, even

though there be slight variations.”81 The prior communications in Foley and Esco’s definition

involve a conscious choice by participating rivals to communicate intentions by means that lack

efficiency justifications. These indicia of culpability distinguish their means of coordination from

simple interdependence in ways courts can identify.82

These examples demonstrate the importance of recognizing and defining the category of

tacit agreement. Mere interdependence is lawful “not because such pricing is desirable (it is not),

but because it is close to impossible to devise a judicially enforceable remedy for ‘interdependent’

pricing. How does one order a firm to set its prices without regard to the likely reactions of its

competitors?”83 A Section 1 agreement, then, should minimally be a form of coordination brought

79 340 F.2d 1000 (9th Cir. 1965).

80 Id. at 1007.

81 Id. at 1008. Cf. AREEDA & HOVENKAMP, supra note 9, ¶ 1418b3, at 118 (describing the series of

statements in Foley as “either (a) a commitment to a common course of action or (b) a prediction of each’s

probable reaction to another’s probable reaction to another’s price increase”; in either case the scenario is

“close enough to a commitment to be deemed a traditional agreement” or at least a “facilitating practice”).

Tacit agreement, I suggest, is more accurate explanation of Foley, and a better basis for understanding Foley

as a precedent, than these characterizations.

82 In the vertical context, the Colgate doctrine permits a firm to announce a “suggested resale price”

without forming an agreement with retailers who comply. United States v. Colgate & Co., 250 U.S. 300,

307 (1919). The buyer’s compliance is, in principle, tacit agreement to the announced resale price condition,

with the same effects as if the buyer and seller had expressly agreed. Leegin Creative Leather Prods., Inc.

v. PSKS, Inc., 551 U.S. 877, 902–03 (2007) (“The economic effects of unilateral and concerted price setting

[under Colgate] are in general the same.”). It is artificially characterized as “unilateral” to protect the seller’s

right to refuse to sell to dealers for its own business reasons, a policy concern that does not apply in the

horizontal context.

83 Clamp-All Corp. v. Cast Iron Soil Pipe Inst., 851 F.2d 478, 484 (1st Cir. 1988).

22

about by means that have no significant efficiency justification (in the sense of an intent to achieve

an efficiency goal) and that courts can practically regulate. An express agreement fits the

description: it represents a “process of reaching supracompetitive marketplace outcomes—what

may be termed the ‘forbidden process’ of negotiation and exchange of assurances. The forbidden

process consists of behavior that can be enjoined.”84 Tacit agreement also satisfies these

conditions. Herbert Hovenkamp argues that whether plus factors suggest agreement is “not so

much whether a ‘meeting of the minds’ in the common-law contract sense existed as whether these

are types of behavior that the law should suppress,” like “communications among rivals about

planned output changes.”85 The idea of tacit agreement captures this last insight that

communications about future competitive choices can bring about a Section 1 agreement. Private

expressions of intention are inefficient actions that may result in noncompetitive equilibria.

Experiments have shown that these sorts of communications can help effectively coordinate

collusive equilibria by limiting the uncertainty facing rivals, without providing any substantial

84 Jonathan B. Baker, Identifying Horizontal Price Fixing in the Electronic Marketplace, 65

ANTITRUST L.J. 41, 48 (1996). Indeed, when it can be proven by direct evidence, an exchange of assurances,

completes a Sherman Act agreement and can be per se unlawful without independent conduct (or any kind

of acts in furtherance of the agreement), even if they do not restrict output in particular cases. United States

v. Socony-Vacuum Oil Co., 310 U.S. 150, 224 n.59 (1940) (observing “that conspiracies under the Sherman

Act are not dependent on any overt act other than the act of conspiring” and that “[i]t is the ‘contract,

combination . . . or conspiracy, in restraint of trade or commerce’ which s 1 of the Act strikes down, whether

the concerted activity be wholly nascent or abortive on the one hand, or successful on the other”) (citations

omitted). Express agreements are considered potentially dangerous, even if they are unenforceable cheap

talk. Joseph Farrell & Matthew Rabin, Cheap Talk, 10 J. ECON. PERSP., Summer 1996, at 103, 107

(“Sometimes there is no incentive to lie, and cheap talk will fully convey private information.”).

85 Herbert J. Hovenkamp, The Pleading Problem in Antitrust Cases and Beyond, 95 IOWA L. REV.

55, 62 (2010). In my account of tacit agreement, the behavior (if not an express agreement) must coordinate

interdependent conduct.

23

benefits to consumers.86 When they do, courts can enjoin or penalize the actions without

unacceptable error costs.87

Twombly insisted on the need to “limit the range of permissible inferences” from parallel,

interdependent conduct. Recognition of tacit agreement requires courts not to limit the range of

permissible inferences to those that permit an inference only of express agreement or a narrowly

understood “traditional conspiracy.” There will undoubtedly be cases in which circumstantial

evidence of communications and subsequent interdependent behavior may permit a reasonable

inference of either an express agreement or a tacit agreement.88 In cases like Foley, however, the

86 On the role of even limited communication in reducing strategic uncertainty and facilitating

coordination, see Miguel A. Fonseca & Hans-Theo Normann, Explicit vs. Tacit Collusion—The Impact of

Communication in Oligopoly Experiments, 56 EUR. ECON. REV. 1759, 1760 (2012) (reporting “clear

evidence that talking leads to higher profits in markets with any number of firms,” with disproportionately

greater benefits in markets with a medium number; the benefits also persist, even after communication

ceases); Andreas Blume & Andreas Ortmann, The Effects of Costless Pre-Play Communication:

Experimental Evidence From Games With Pareto-Ranked Equilibria, 132 J. ECON. THEORY 274, 288

(2007) (“[W]ith repeated interaction costless messages [of the sender’s intent to play an equilibrium]

preceding games with Pareto-ranked equilibria can dramatically facilitate participants’ coordination on the

Pareto-dominant equilibrium [and] help overcome well-documented problems of strategic uncertainty,

equilibrium selection, and coordination failure.”); Anthony Burton & Martin Sefton, Risk, Pre-Play

Communication and Equilibrium, 46 GAMES & ECON. BEHAVIOR 23, 35 (2004) (finding that under strategic

uncertainty, “equilibrium is much more likely to obtain when subjects have opportunities to send messages

to one another prior to making choices”); Michihiro Kandori & Hitoshi Matsushima, Private Observation,

Communication and Collusion, 66 ECONOMETRICA 627, 647 (1998) (“[C]ommunication is an important

means to resolve possible confusion among players in the course of collusion during repeated play.”).

87 Where the evidence shows that rivals continue to act independently, there is no tacit agreement.

See, e.g., In re Dairy Farmers of Am., Inc. Cheese Antitrust Litig., 801 F.3d 758, 763 (7th Cir. 2015)

(finding that, despite communications among defendants, there was “voluminous evidence” that the

defendant’s purchases on a commodities exchange “were the result of its own interest in restoring a certain

spread between the prices of block and barrel cheese” precluded inference of agreement).

88 For example, participation in trade association meetings, with the attendant opportunity to

communicate about prices, is generally an insufficient basis by itself for inferring that a subsequent parallel

price increase was collusive. In re Graphics Processing Units Antitrust Litig., 527 F. Supp. 2d 1011, 1023

(N.D. Cal. 2007) (“Attendance at industry trade shows and events is presumed legitimate . . . .”). But if

there are direct pricing discussions at those meetings followed by an unusual pattern of price increases, the

inference of agreement, either tacit or express, is stronger. See, e.g., In re Pressure Sensitive Labelstock

Antitrust Litig., 566 F. Supp. 2d 363, 372 (M.D. Pa. 2008) (finding sufficient inference of agreement where

firms allegedly held “direct discussions about the need to collaborate on price increases during” a trade

association meeting and, one month later, a firm that was usually a price follower led a price increase).

24

evidence of communications and interdependent action may make it reasonable for a fact-finder

to infer only that rivals had communicated their intentions sufficiently to coordinate a tacit

agreement. In that range of cases, recognizing tacit agreement (appropriately defined) does matter.

The definition is also important in cases in which there is evidence of repeated communications,

but with few specifics about their content. In cases like these, a more inclusive definition of

agreement can determine whether the evidence communication and coordination is sufficient to

raise a reasonable inference of agreement.

C. COMMUNICATIONS THAT COORDINATE INTERDEPENDENT CONDUCT

The legal treatment of oligopolistic coordination depends in large part on the

communications the rivals use as means of coordination. As the rest of this article will suggest, the

two most critical variables in identifying agreement in these circumstances are the audiences and

the subject matter of the communications. The court is more likely to find an agreement where the

communications are private rather than public and where they relate to future rather than present

competitive conditions. If a communication is private—solely among the rivals—it is less likely

to benefit consumers than a public announcement, and more likely to be surreptitious and focused

on the concerns of the rivals themselves.89 As one court put it, a “direct and secret price discussion

between competitors is more probative of a conspiracy than are indirect and public

89 Carlton et al., supra note 6, at 432 (observing that “whether or not the communication is public

is important because consumers may benefit from the information externalized,” but also noting that

“private exchange of information is not certain to be anticompetitive and, furthermore, consumers may be

uninterested in the information”). See also In re Ethylene Propylene Diene Monomer (EPDM) Antitrust

Litig., 681 F. Supp. 2d 141, 176 (D. Conn. 2009) (finding “evidence of the frequent and friendly

communications between the defendants and the secrecy of their meetings” raised an inference of

agreement); Standard Iron Works v. ArcelorMittal, 639 F. Supp. 2d 877, 894 (N.D. Ill. 2009) (“Public

statements about output reduction, in the form of press releases or SEC filings, are fundamentally distinct

from statements made at trade meetings directly to competing executives, and Defendants proffered non-

collusive purpose for the former does not, in all instances, account for the latter.”).

25

communications, ostensibly undertaken by the conspiring competitors to ‘signal’ to one another.”90

Communications by individuals with pricing or other relevant decisionmaking authority are, of

course, most probative.91 And the communication (again, particularly by upper management) is

about future competitive conditions it is more likely to be aimed at limiting strategic uncertainty

and thus contributing to the coordination of competitive behavior. If the communications that

coordinate interdependent conduct share both of these characteristics, they are almost certain to

amount to an agreement.

By contrast, if firms communicate publicly with both consumers and competitors, the

communication is very likely to promote efficient market functions by enhancing consumer choice

and is less likely to be an effective means of coordination. If they communicate only about present

pricing (or some other competitive variable), they will be less able to coordinate their strategic

choices. If communications are both fully public and relate only to present pricing, courts are

certain to treat the rivals’ conduct as independent or at most simply interdependent pricing.

The categories of communications can be arranged in a grid with cells corresponding to

the four combinations of audience and temporal focus:

90 In re Polyurethane Foam Antitrust Litig., No. 1:10 MD 2196, 2015 WL 520930, at *10 (N.D.

Ohio Feb. 9, 2015). See also AREEDA & HOVENKAMP, supra note 9, ¶ 1435c, at 255-56 (observing that

public speech can coordinate competitive behavior, but often has plausible benefits).

91 See In re Flat Glass Antitrust Litig., 385 F.3d 350, 358–59 (3d Cir. 2004) (observing that “price

discussion among low level sales people has little probative weight,” but that it is a “far different situation

where upper level executives have secret conversations about price”); In re Baby Food Antitrust Litig., 166

F.3d 112, 124–26 (3rd Cir. 1999) (“Evidence of sporadic exchanges of shop talk among field sales

representatives who lack pricing authority is insufficient to survive summary judgment.”); In re

Polyurethane Foam Antitrust Litig., No. 1:10 MD 2196, 2015 WL 520930, at *21 (N.D. Ohio Feb. 9, 2015)

(finding “substantial evidence of senior Defendant employees engaging in price discussions, sometimes

providing one another advance notice of [price increase announcements or PIAs], sometimes sharing PIA

information during the implementation period when a proposed price increase was being tested in the

market, and sometimes discussing the progress of past PIAs”); In re Currency Conversion Fee Antitrust

Litig., 773 F. Supp. 2d 351, 370 (S.D.N.Y. 2011) (“[H]ere, senior in-house counsel attended the May 25

Meeting and engaged in discussions on a far higher plane than mere ‘shop talk’ among low-level

employees.”).

26

Temporal Focus

Future Present A

udie

nce

Private 1

2

Public 3 4

The numbers in the cells correspond to the relative probative value of the communications in the

proof of agreement, other things equal, with one being the most probative and four the least.92

Private communications about future competitive choices concerning a factor salient to consumers,

particularly prices, are most likely to justify an inference of agreement.93 Foley is a paradigmatic

instance that I will discuss further in the next Part, along with other examples. I will also discuss

a related pattern of communications, in which communications by upstream or downstream firms

with the rivals (rather than communications among the rivals themselves) coordinate a tacit

horizontal agreement. Private communications about present prices, which I discuss in Part III, are

less probative of per se illegal agreement to fix prices, but may permit an inference of an agreement

to exchange prices, which the court will evaluate under the Rule of Reason.

92 The probative value of any piece of evidence varies depending upon other evidence of the

circumstances. In re High Fructose Corn Syrup Antitrust Litig., 295 F.3d 651, 655 (7th Cir. 2002) (warning

of the “trap” in evaluating evidence of conspiracy “to suppose that if no single item of evidence presented

by the plaintiff points unequivocally to conspiracy, the evidence as a whole cannot defeat summary

judgment,” because “evidence can be susceptible of different interpretations, only one of which supports

the party sponsoring it, without being wholly devoid of probative value for that party”).

93 See Kai-Uwe Kühn et al., Fighting Collusion by Regulating Communication between Firms, 16

ECON. POL’Y 167, 186 (2001) (“Our analysis of theory, experimental evidence and case material suggests

that any private communication between firms about planned future conduct including suggestions for

future conduct, even unilateral communication, should be considered a competition reducing restriction in

the sense of Art. 81(1) of the Treaty of Rome.”). To complete a tacit agreement under the Sherman Act

under the proposal in the following pages, the communications must later lead to coordinated oligopolistic

behavior.

27

Public communications about present prices, like the price announcements by service

stations, are not probative of tacit agreement, even if they succeed in coordinating noncompetitive

pricing. As I show in Part IV, courts view them as inseparable from simple interdependent pricing,

unless they include some encoded private message. I show in Part V that public communications

about future prices (or other competitive choices) are usually thought to benefit consumers, and so

are not a basis for inferring tacit agreement, unless the communication is functionally private

because of its context or mechanism and is designed to coordinate competitive behavior.94

Courts sometimes interpret communications as signals conveying something more than

their literal or superficial meaning. Firms may use signals, in this sense, to coordinate

interdependent conduct and even tacit agreement. In earlier work, I have distinguished

conventional signals, prearranged signals, and implicit signals.95 Conventional signals, as I have

already suggested,96 are arbitrary actions, like nodding, that a members of a culture recognize as

having a specific meaning;97 prearranged signals are outwardly benign actions, statements, or

events that parties have agreed to treat as having special significance for them.98 Both conventional

94 Id. at 186 (“[W]e should consider any communication private if there are no plausible direct

positive effects on consumers.”).

95 William H. Page, Signaling and Agreement Under Section 1 of the Sherman Act, 5

CONCURRENCES: COMPETITION L.J., No. 3 (2015).

96 See supra, Part II.B.

97 See, e.g., Blomkest Fertilizer, Inc. v. Potash Corp. of Sask., 176 F.3d 1055, 1063 (8th Cir. 1999)

(“The law forbids nonverbal agreements in restraint of trade, as well as express ones; otherwise, the law

would be emasculated as competitors accomplished the forbidden goal with a wink and a nod.”), superseded

en banc on other grounds, 203 F.3d 1028, 1037 (8th Cir. 2000).

98 For example, in Eastern States Retail Lumber Dealers Ass’n v. United States, 234 U.S. 600, 608–

09 (1914), the government contended that “the purpose in the predetermined and periodical circulation of

[blacklists of offending wholesalers was] to put the ban upon wholesale dealers whose names appear in the

list.” An association circulated a list of wholesalers that were selling directly to consumers. These so-called

blacklists predictably led retailers to refuse to buy from listed suppliers. The government conceded that

“there is no agreement among the retailers to refrain from dealing with listed wholesalers,” but argued that

“he is blind indeed who does not see the purpose in the predetermined and periodical circulation of this

report to put the ban upon wholesale dealers whose names appear in the list.” Id. The Supreme Court

28

and prearranged signals are mainly means of forming or implementing an express agreement

essentially in the same way as words, but with the advantage greater secrecy.99 More important for

the present analysis of tacit agreement are implicit signals—actions or words with general, often

benign meanings for more than one audience, but that also include information that conveys a

special meaning for rivals. These sorts of signals will be most significant in public communications

about future competitive conduct.100

II. PRIVATE COMMUNICATIONS ABOUT FUTURE ACTIONS

In the last Part, we saw that tacit agreement means interdependent actions coordinated by

prior communications of competitive intentions, in circumstances in which the communications

do not provide any benefit to other audiences, especially consumers. In the paradigmatic case,

rivals privately and directly state pricing intentions, then act consistently with the stated intentions.

In this Part, I examine Foley and other historical and recent examples, under different states of

evidence. I also discuss a variant that courts also view as involving tacit agreement—the so-called

“hub-and-spokes” agreement—in which rivals tacitly accede to a proposal, usually from upstream

or downstream firms.

A. COMMUNICATIONS AMONG RIVALS

Communications that convey pricing intentions are most probative of agreement generally;

if the communications have no procompetitive justification, and have “an impact on

concluded that “when . . . by concerted action the names of wholesalers who were reported as having made

sales to consumers were periodically reported to the other members of the associations, the conspiracy to

accomplish that which was the natural consequence of such action may be readily inferred.” Id. at 612. In

our terms, the association prearranged the use of a blacklist to signal which suppliers to boycott.

99 See, e.g., R. Austin Smith, The Incredible Electrical Conspiracy (Part II), FORTUNE, May 1961,

at 161, 164, 210 (describing the electrical equipment conspirators’ use of the phases of the moon as a

mechanism for rotation of winning bidders in public auctions).

100 See infra, Part V.

29

[interdependent] pricing decisions,”101 they justify an inference of a tacit agreement. Substantial

private communications among rivals about their future competitive choices, are most likely to be

(or to allow the inference of) nakedly causal communications or “conversations implying that later

uniformity might prove desirable.”102 In United States v. Foley,103 which I quoted in the last Part,

warrants closer examination on this issue. The court there affirmed criminal price fixing

convictions, despite the absence of evidence of a verbal agreement. The defendants were realtors

who wanted to increase their commission rate from 6 to 7 percent in a depressed housing market,

but “were concededly afraid to undertake such a move for fear that they would be unable

successfully to compete with firms still at six percent.”104 One had already attempted such an

increase and withdrew it when others did not follow.105 At this point, one of the brokers named

Foley called a meeting, to be held over dinner:

At the dinner Foley rose, made some prefatory remarks and then stated that his firm

was in dire financial condition. Saying that he did not care what the others did, he

then announced that his firm was changing its commission rate from six percent to

seven percent. Testimony as to what was said by various persons in the ensuing

discussion is greatly in conflict, but there was evidence from which the jury could

find that each of the individual defendants and a representative of each corporate

defendant not represented by one of the individual defendants expressed an

intention or gave the impression that his firm would adopt a similar change. The

101 In re Baby Food Antitrust Litig., 166 F.3d 112, 125 (3d Cir. 1999).

102 White v. R.M. Packer Co., 635 F.3d 571, 576 (1st Cir. 2011) (quoting Brown v. Pro Football,

Inc., 518 U.S. 231, 241 (1996)).

103 598 F.2d 1323 (4th Cir. 1979).

104 Id. at 1331.

105 Id.

30

discussion also included reference to the earlier unsuccessful effort by [one realtor]

to adopt a seven percent policy, from which the jury could conclude that defendants

knew that their cooperation was essential.106

There was evidence that, following the meeting, “each defendant did in fact begin to take

substantial numbers of seven percent listings” and in some instances rivals tried “to hold their

fellows to the ‘agreement.’”107

The guests at that dinner were careful to avoid an express verbal agreement. They did not

make promises and limited themselves to statements or less direct indication of intention; they

knew, based on experience and on what they said at the meeting, that they had to stand or fall

together; and they then tacitly completed the agreement by acting consistently with those

statements, albeit in some instances after some further coaxing. The statements had no credible

efficiency justification or purpose; they conveyed no information to consumers or other audiences,

because only competitors were present; their naked purpose was to permit rivals to coordinate

future conduct. In these circumstances, the court properly concluded that a jury could reasonably

106 Id. at 1332.

107 Id. See also Esco Corp. v. United States, 340 F.2d 1000, 1007 (9th Cir. 1965) (hypothesizing a

meeting of four rivals at which three state that they plan to set a price of X dollars; all four then set just that

price). Other cases cite Foley or Esco or both for this idea of agreement. See Standard Iron Works v.

ArcelorMittal, 639 F. Supp. 2d 877, 895–96 (N.D. Ill. 2009) (denying a motion to dismiss, where “trade

meeting statements made directly to competing executives [that] encourage[ed] industry production

restraint . . . were followed closely by unprecedented industry-wide production cuts, exactly as encouraged

and represented by Defendant executives.”); In re TFT–LCD (Flat Panel) Antitrust Litig., 586 F. Supp. 2d

1109, 1116 (N.D. Cal. 2008) (“Courts have held that a conspiracy to fix prices can be inferred from an

invitation followed by responsive assurances and conduct.”); Wall Prods. Co. v. Nat’l Gypsum Co., 326 F.

Supp. 295, 324–25 (N.D. Cal. 1971) (“It was enough that, knowing that concerted action was contemplated

and invited, the distributors gave their adherence to the scheme and participated in it. Each distributor was

advised that the others were asked to participate; each knew that cooperation was essential to successful

operation of the plan.”).

31

infer that the subsequent interdependent moves to the 7 percent rate completed a Section 1

agreement.108

Another paradigmatic instance of this sort of arrangement was the Gary Dinners system,

conceived and supervised by Judge Elbert Gary, the chairman of U.S. Steel, in the early Twentieth

Century.109 Formed by merger in 1901, U.S. Steel was a holding company composed of twelve

operating companies,110 some of which had participated in highly organized pools and cartels for

many years before the merger.111 In an effort to comply formally with the antitrust laws,112 Judge

108 See also Havens v. Maritime Commc’ns/Land Mobile, LLC, No. 11-993, 2014 WL 4352300

(D.N.J. Sept. 2, 2014) (distinguishing Foley on the ground that the defendants, unlike the brokers in Foley,

did not act consistently with the alleged statements). But see Market Force, Inc. v. Wauwatosa Realty Co.,

906 F.2d 1167, 1172–73 (7th Cir. 1990), which characterized as mere “conscious parallelism” a scenario

in which (1) a real estate agency mailed rivals “copies of its policy expressing its intention to pay 20% of

the selling agent’s commission to buyers’ brokers” and (2) the rivals “subsequently adopted similar

policies.” The fact that only a single rival announced its intentions to the others was enough to distinguish

Foley, but the court went further. In a footnote, it said Foley did “not rely on the dinner party and

announcement to find a conspiracy”; it only affirmed the convictions because there was evidence of later

“letters and phone calls among the realtors exhorting each other to charge the higher rate and fearing that

deviation would ruin the plan.” Id. at 1172 n.7. Certainly the court in Foley did not rely solely on Foley’s

statement the dinner party; all of the rivals’ statements then and their subsequent, eventual consistent price

increases were also essential. Foley did point to evidence of later “instances in which members of the

conspiracy sought after the September 5 dinner to hold their fellows to the ‘agreement,’” 598 F.2d at 1332,

but it did not say those kinds of communications would have been necessary to a finding of agreement if

the rivals had acted consistently with the statements at the meeting without any further communications.

The inference of tacit agreement would have been equally strong, if not stronger, had the rivals, after the

meeting, perfectly followed their stated intentions without cajolery. The later communications were

significant in Foley mainly because some of the rivals deviated and needed some prodding. The court

affirmed the conviction of one of the defendants based solely on his statements at the meeting and

subsequent consistent pricing. Id. at 1334.

109 See generally William H. Page, The Gary Dinners and the Meaning of Concerted Action, 62

SMU L. REV. 597 (2009).

110 Donald O. Parsons & Edward John Ray, The United States Steel Consolidation: The Creation

of Market Control, 18 J.L. & ECON. 181, 182 (1975).

111 Transcript of Record on Appeal from the District of New Jersey, vol 2, at 816–24 (Testimony

of John C. Langan), United States v. U.S. Steel Corp., 251 U.S. 417 (1920) [hereinafter Transcript of

Record], available at The Making of Modern Law: U.S. Supreme Court Records and Briefs, 1832-1978,

galenet.galegroup.com/servlet/SCRB?finalAuth=true.

112 See also UNITED STATES BUREAU OF CORPORATIONS, TRUST LAWS AND UNFAIR COMPETITION

15 (1916) (observing that, because “[l]egal difficulties had at an earlier date . . . discouraged the use of

written price and pooling agreements as means of combination,” firms adopted the form of the “gentlemen’s

32

Gary ordered U.S. Steel’s constituent corporations not to participate in the usual forms of price

fixing.113 At the same time, Gary adopted what he believed to be a lawful but effective alternate

way of limiting competition: dinners and meetings of producers of various lines of fabricated steel

at which the rivals would, by turns, state the prices they intended to charge. Participants were to

make no promises,114 but would adhere to the announced prices until they decided to change them,

in which case they would notify their rivals.115 In the course of deciding the government’s

monopolization suit against U.S. Steel in 1915, the district court held that, by the dinners and

meetings, U.S. Steel and its rivals had achieved “action with a common object” and “an

understanding concerning the maintenance of price”116 that was “equivalent to an agreement.”117

agreement,” which aimed to “procure substantial harmony in policy among competitors regarding volume

of output and prices, without specific written or oral agreements, but rather by tacit understandings and the

communication of information regarding the production and price of the various cooperating interests in

order to insure good faith”).

113 Brief of Appellee, filed Mar. 3, 1917, at 206–07, United States v. U.S. Steel Corp., 251 U.S. 417

(1920).

114 Transcript of Record, supra note 111, vol. 11, at 4196 (Charles M. Schwab); id. at vol. 5 at 1777

(Crawford); id. vol. 5 at 1836–37 (Test. of James Anson Campbell) (stating that, as chairman, he would

“always [make] the statement . . . that there could be no agreements among us” but that he “called on each

[member] to state what their policy would be in future with reference to the sale of their products and with

reference to price”; there would usually, though not always, be a “general understanding; that is, we had

the statement of the different representatives that their policy would be to market their product at the then

prevailing price until they notified their competitors that they wanted to change their price”).

115 Id. vol. 6, at 2104 (Willis L. King) (stating that he “would undoubtedly have felt that I should

notify [rivals] if I cut the price,” but only out of common decency, not because of an agreement).

116 United States v. U.S. Steel Corp., 223 F. 55, 159 (D.N.J. 1915), aff’d, 251 U.S. 417, 444–45

(1920).

117 Id. at 160 (emphasis added). In affirming on direct review after World War I, the Supreme Court

agreed that the various means of price control, including the “social form of dinners,” were unlawful. United

States v. U.S. Steel Corp., 251 U.S. 417, 440 (1920). It concluded, however, it was “unable to see that the

public interest will be served” by dissolving the company. Id. at 457.

Interestingly, in the market for Bessemer steel rails, fabricators maintained a price of $28 for years

without any meetings or communications comparable to the Gary Dinners system. U.S. Steel, 224 F. at 154.

The district court found that the price was “tacitly accepted and continued by the sales managers of different

rail companies” who “simply followed that basic price to prevent the ruinous rail wars of the past.” Id.

Without the prior coordinating communications, this sort of tacit collusion was not an agreement.

33

In earlier work, I discuss the Gary Dinners as “concerted action,”118 but they are also

appropriately classified as a tacit agreement, as defined in the last Part: interdependent conduct

coordinated by causal, direct prior communications. Judge Gary thought his arrangement would

escape the Sherman Act, because the rivals did not form express verbal agreements; they only

innocently announced their intent about pricing, then acted honorably by pricing consistently with

their statements. But the courts properly recognized that the announcements functioned as

proposals to which the rivals in each product line tacitly agreed by adopting noncompetitive prices.

In cases like Foley and the Gary Dinners, the words of the rivals communicated more than their

literal meaning. When a firm tells its rivals that it plans to raise prices, it is stating its own intention

and simultaneously “implying that later uniformity might prove desirable.”

In both Foley and U.S. Steel the rivals met to share their pricing intentions,119 but private

communications need not involve a physical meeting. The same result should follow where there

118 Page, Gary Dinners, supra note 109, at 614–18.

119 Another example is United States v. Champion International Corp., Cr. No. 74-183, 1975 WL

920 (D. Ore. July 16, 1975), in which timber company representatives met to discuss future auctions of

timber; at the meetings each “told the others in which upcoming sale or sales his company was interested.

Sometimes one of them would go into the reasons for this interest.” Id. at *2. “No one really committed

himself not to bid but in sale after sale over a four year period, the one who had expressed the highest

interest in a sale was the one who took the sale without opposition.” Id. at *3. The court of appeals affirmed

criminal convictions, 557 F.2d 1270 (9th Cir. 1977), observing that:

In a general way, the extent of an individual operator’s interest in a future sale could have

been predicted by anyone familiar with the operator’s hauling distances, his product mix,

his manufacturing capacity, and the other factors that determine a sale’s relative desirability

to that operator. However, the defendants did not leave the exchange of this information to

chance. During the time covered by the indictment the defendants advised each other about

the future sales upon which they were most likely to bid. Whether or not anyone ever

agreed at those meetings to bid or to refrain from bidding in any way, there was no doubt

that the defendants “had an understanding” about bidding.

Id. at 1273. The court agreed with the government that, even though there was no “direct evidence of an

express agreement, . . . the circumstantial evidence proved the existence of the tacit agreement found by

the judge.” Id. In this scenario, the parties stated their intention by their expressions of the intensity of their

interest, then completed the tacit agreement by acting consistently with those statements. Although, as the

court of appeals observed, the parties might have been able to coordinate similar auction outcomes based

34

is circumstantial evidence of direct written communications,120 particularly of intention121 and the

rivals act interdependently and consistently with these sorts of communications. Consider the

court’s summary of the evidence of communications in Flat Glass:

[T]he exchanges of information are . . . tightly linked with concerted behavior and

therefore they appear more purposive. Several of the key documents emphasize that

the relevant price increases were not economically justified or supportable, but

required competitors to hold the line. Others suggest not just foreknowledge of a

single competitor’s pricing plans, but of the plans of multiple competitors.

Predictions of price behavior were followed by actual price changes.122

on public information about rivals’ likely interest, the private exchanges of information reduced the

uncertainty of the coordination process.

Areeda and Hovenkamp argue that “[i]n Champion, we are entitled to find a bid rotation conspiracy

event though we do not know what assurances, if any, the parties exchanged at the meetings [because]

permitting parties to meet under the circumstances and for the discussions at issue should be discouraged,

given the inherent dangers to competition and the lack of any obvious social virtue.” Cf. AREEDA &

HOVENKAMP, supra note 9, ¶ 1418b4, at 120. This article’s proposed definition of tacit agreement supports

the convictions in Champion more directly: the rivals need not have exchanged assurances at all, because

they stated their intentions in private communications that resulted in a consistent pattern of bid rotation.

120 See, e.g., E.W. French & Sons, Inc. v. Gen. Portland, Inc., No. 94-55525, 1995 WL 470825, at

*4 (9th Cir. Aug. 8, 1995) (finding evidence that a rival maintained a “plain white envelope” mailing list

to exchange price lists with rivals before distributing the prices to consumers raised an inference of price

fixing); In re Foreign Exch. Benchmark Rates Antitrust Litig., 74 F. Supp. 3d 581, 591–92 (S.D.N.Y. 2015)

(foreign exchange traders “used various electronic communications platforms, particularly chat rooms and

instant messaging, to share ‘market-sensitive information with rivals’” before setting the published closing

exchange rate); Sun Microsystems, Inc. v. Hynix Semiconductor, Inc., 622 F. Supp. 2d 890, 904–07 (N.D.

Cal. 2009) (finding frequent exchanges of detailed pricing information including “price targets” and “price

forecasts” and “competitive feedback” raised an inference agreement).

121 Courts emphasize the importance of communications about future pricing. See, e.g., In re Flash

Memory Antitrust Litig., 643 F. Supp. 2d 1133, 1143–44 (N.D. Calif. 2009) (identifying communications

about planned price increases and the need not to reduce prices); In re Medical X-Ray Film Antitrust Litig.,

946 F. Supp. 209, 218–21 (E.D.N.Y. 1996) (holding that evidence that rivals exchanged internal

memoranda concerning plans to increase prices and otherwise learned privately of upcoming price increases

supported an inference of agreement).

122 In re Flat Glass Antitrust Litig., 385 F.3d 350, 369 (3d Cir. 2004). See also In re Static Random

Access Memory (SRAM) Antitrust Litig., No. 07–md–01819 CW, 2010 WL 5138859, at *8 (N.D. Cal.

Dec. 10, 2010) (describing evidence of “exchange of critical business information, including price” with

35

Here, the court identifies all of the critical elements of tacit agreement, although it did not use that

term: the communication of pricing intent; interdependent pricing; and a basis for an inference of

a causal relationship. Communications through an intermediary, although more cumbersome, are

also private and can serve the same function as direct communication.123

In some instances, the court may infer the existence of the requisite communications from

other evidence or allegations.124 In one case, the complaint alleged that an employee of one

manufacturer sent an email to a rival’s employee asking: “Are you willing to exchange product

roadmaps again?”125 The firms argued that the email suggested that they had no agreement to share

confidential information, but the court found it plausible, on a motion to dismiss, to infer “that the

companies had an agreement to exchange the information from time to time and [the sender] was

some rivals that allowed rivals to better control supply and could lead to a finding of collusion by a trier of

fact).

123 In In re Domestic Drywall Antitrust Litigation, M.D.L. No. 2437, 13-MD-2437, slip op. at 118

& 131 (E.D. Pa. Feb. 18, 2016), the court held evidence that firms provided information to research analysts

knowing that the analysts would share the information with rivals suggested that the firms “used the

research analysts as a conduit to signal to the other manufacturers during the class period,” although it was

“insufficient to permit the inference that Defendants actually reached an agreement by communicating

through analysts.” Slip op. at 118–19. The Drywall court cited In re Insurance Brokerage Antitrust

Litigation, 618 F.3d 300, 337 (3d Cir. 2010); In re Titanium Dioxide Antitrust Litigation, 959 F. Supp. 2d

799, 829 (D. Md. 2013); and In re Polyurethane Foam Antitrust Litigation, No. 1:10 MD 2196, 2015 WL

520930, at *41 (N.D. Ohio Feb. 9, 2015), all of which suggest that communications through an intermediary

“could be indicia of a price fixing agreement.” Slip op. at 122.

124 Twombly itself suggested that, in “a traditionally unregulated industry with low barriers to entry,

sparse competition among large firms dominating separate geographical segments of the market could very

well signify illegal agreement.” Bell Atl. Corp. v. Twombly, 550 U.S. 544, 567 (2007). See also In re Potash

Antitrust Litig., 667 F. Supp. 2d 907, 937 (N.D. Ill. 2009) (finding plaintiffs’ allegations sufficient to

withstand a motion to dismiss where they alleged “parallel production cuts which at least for some

Defendants represented a ‘radical change’ in behavior; overlapping business ventures; specific meetings

attended by Defendants which precipitated production cuts and price increases; and a ‘joint sales

suspension’ that was contrary to Defendants’ economic interests”), aff’d sub nom. Minn-Chem, Inc. v.

Agrium, Inc., 683 F.3d 845 (7th Cir. 2012). Sham or “complementary” bidding is often mentioned as an

example of strong circumstantial evidence of express agreement. AREEDA & HOVENKAMP, supra note 9, ¶

1420b, at 154 (“A strong inference of coordinated behavior arises when a participant actively seeks to lose

a bid.”).

125 In re Static Random Access Memory (SRAM) Antitrust Litig., 580 F. Supp. 2d 896, 901 (N.D.

Cal. 2008).

36

inquiring whether it was time for the next exchange.”126 Internal emails discussing “information

obtained from other Defendants” suggested “that Defendants were aware that the purpose of

sharing this information was to ‘stabilize or raise the price of SRAM sold in the United States and

elsewhere.’”127 The term “roadmap” implied that the information related to future prices and

demand conditions.

An inference of tacit agreement is, of course, most reasonable when the communications

have to do with terms central to competition, particularly price or output.128 The correlation

between present communication among rivals about their future behavior and interdependence is

less clear-cut if the communications relate to a variable less salient to consumer choice and more

likely to have plausible independent motivations.129 For example, one court recently concluded

after a bench trial that banks’ parallel adoptions of class-action barring arbitration clauses were

not interdependent, despite extensive evidence of dozens of meetings of counsel over several years

in which they shared information, “tried to determine their competitors’ plans and experiences

regarding arbitration,” and pursued “collective efforts to establish [the clauses] as an industry

norm.”130 Although these communications undoubtedly formed a benign (in the court’s view)

126 Id. (citing United States v. Container Corp. of Am., 393 U.S. 333, 335 (1969)).

127 Id. at 902 (citing Container, 393 U.S. at 335). See also id. at 898, 903 (observing that, because

of its concentration and product homogeneity, “the SRAM market was one in which such information

exchanges would lead to price stabilization or increases.”)

128 See, e.g., Toys “R” Us, Inc. v. FTC, 221 F.3d 928, 936 (7th Cir. 2000) (finding sufficient

evidence, under Interstate Circuit, that Toys “R” Us had induced toy manufacturers to boycott retail

warehouse clubs; acceding to the dominant retailer’s demand, the manufacturers acted interdependently,

because each “was afraid to curb its sales to the warehouse clubs alone, because it was afraid its rivals

would cheat and gain a special advantage in that popular new market niche”).

129 See, e.g., Burtch v. Millberg Factors, Inc., 662 F.3d 212, 223 (3d Cir. 2011) (exchanging even

“forward-looking” credit information about customers, unlike price information, does not suggest

conspiracy, because the information “protect[s] competitors from insolvent customers”).

130 Ross v. Am. Express Co., 35 F. Supp. 3d 407, 452–53 (S.D.N.Y. 2014) (“[T]he final decision

to adopt class-action-barring arbitration clauses was something the Issuing Banks hashed out individually

and internally.”), aff’d sub nom. Ross v. Citigroup, Inc., 630 F. App’x 79 (2d Cir. 2015).

37

agreement to share information and “to establish class-action-barring arbitration as an industry

norm,” the court was unwilling to infer “a separate, illegal agreement to collusively adopt and

maintain class-action-barring arbitration clauses,”131 because “avoiding class actions through

arbitration was in each Issuing Banks’ independent self interest, regardless of whether its

competitors also adopted such a provision.”132 This reasoning assumes that cardholders do not care

whether they have a right to sue their issuer in a class action, so the issuer can safely include a

waiver provision in their card agreement regardless of whether its rivals do so.

B. VERTICAL COMMUNICATIONS THAT COORDINATE HORIZONTAL

INTERDEPENDENCE: HUB-AND-SPOKES AGREEMENTS

In Interstate Circuit,133 a representative of related circuits of first-run movie exhibitors

wrote a letter to movie distributors, naming all of them as addressees, and demanding that they

require second-run exhibitors to raise their admission prices and stop showing double features.134

131 Id. at 452. The inference of an agreement to exchange present pricing information is discussed

in Part IV.B. below.

132 Id. at 453. The court found that these legitimate “communications resembled those of trade

associations or lobbying groups.” Id. at 452. The court went on to say in dicta (to aid appellate review, as

the court put it) that, had the plaintiffs carried their burden on the issue of agreement, the court would have

concluded under a quick-look version of the rule of reason that the agreement was a violation of Section 1.

Id. at 453–56. Because the court of appeals affirmed the principal holding, it did not reach this issue. Ross,

630 F. App’x at 82 n.4.

133 Interstate Circuit, Inc. v. United States, 306 U.S. 208 (1939). Both of the cases quoted in

Twombly that refer to tacit agreements cited Interstate Circuit as an illustration. Theatre Enterprises cited

the case alongside American Tobacco as one in which the courts properly inferred an agreement from

parallel business conduct. Theater Enters. v. Paramount Film Distrib. Corp., 346 U.S. 537, 540 (1954).

White cited only Interstate Circuit as an illustration of tacit agreement, calling it a “seminal case.” White

v. R.M. Packer Co., 635 F.3d 571, 576 (1st Cir. 2011). For other cases linking Interstate Circuit to tacit

agreement, see DM Research, Inc. v. College of American Pathologists, 170 F.3d 53, 56 (1st Cir. 1999);

Barry v. Blue Cross of California, 805 F.2d 866, 869 (9th Cir. 1986); Opticon, Inc. v. Zazzu, No. 94 CR

6256, 1995 WL 67656 at *1 n.2 (S.D.N.Y. Feb. 17, 1995), and In re Mid-Atlantic Toyota Antitrust

Litigation, 560 F. Supp. 760, 772–73 (D. Md. 1983).

134 Interstate Circuit, 306 U.S. at 221 (the trial court properly inferred an express “agreement from

the nature of the proposals made on behalf of Interstate and Consolidated; from the manner in which they

were made; from the substantial unanimity of action taken upon them by the distributors; and from the fact

that appellants did not call as witnesses any of the superior officials who negotiated the contracts with

38

The distributors largely acceded to the demands of one of the circuits and uniformly rejected the

demands of the other circuit.135 The Supreme Court found it proper to infer that the distributors

had expressly conspired, even though there was no testimony describing direct communications

among them.136 More important for our purposes, the Court held that, even if distributors did not

communicate directly, their uniform, interdependent actions in response to the exhibitors’

invitation to participate in the arrangement was sufficient to satisfy the agreement requirement:

“It was enough that, knowing that concerted action was contemplated and invited, the distributors

gave their adherence to the scheme and participated in it. Each distributor was advised that the

others were asked to participate; each knew that cooperation was essential to successful operation

of the plan.”137

The Court later described this scenario as “a tacit agreement to restrain competition

between the distributors.”138 The distributors, confronted by the first-run exhibitors’ ultimatum,

Interstate or any official who, in the normal course of business, would have had knowledge of the existence

or non-existence of such an agreement among the distributors”)

135 Id. at 218–19.

136 The form of the letter assured that each distributor knew that all of the others had received the

same proposal, and there was “strong motive for such unanimity of action.” Id. at 225. The pattern of

uniform acceptance and rejection of the proposals to change established marketing practices supported the

inference of an express agreement, particularly “when uncontradicted and with no more explanation than

the record affords.” Id. at 225. The distributors offered only testimony from lower-level managers, not ones

with decision-making authority: “The production of weak evidence when strong is available can lead only

to the conclusion that the strong would have been adverse.” Id. at 226.

137 Interstate Circuit, 306 U.S. at 226. Cf. Shamir & Shamir, supra note 30, at 656 (discussing the

possibility of inferring collusion where “the private information of one retailer, which is shared with the

manufacturer, [is] be leaked unintentionally to the retailer’s competition through the pricing mechanism.”).

The authors suggest that “when the manufacturer determines the price based on the private information he

receives from the retailer, a competing retailer can observe this price and infer the private information of

the competitor from this price; since the price of the manufacturer is a function of the private information

of the retailer, observing the manufacturer’s price can be equivalent to observing the private information

itself.” Id. In the text, I am concerned with the inverse case of a proposal from a common source to upstream

firms as a means of forming a tacit agreement among the recipients of the information.

138 First Nat’l Bank of Ariz. v. Cities Serv. Co., 391 U.S. 253, 287 (1968).

39

“all had the same motive to enter into a tacit agreement [that] would enable them to increase their

royalties by forcing a rise in admission prices without the danger of competitors enlarging their

share of the subsequent-run market by refusing to impose similar restrictions.”139 Similarly, the

First Circuit in White described Interstate Circuit as an exemplar of tacit agreement, because:

the distributors, who never communicated directly with one another, nonetheless

had entered into a tacit agreement with one another by acting in accordance with

the letter’s demands, because the letter made it clear that all eight had received the

letter, the economic context made it clear that all eight needed to act uniformly or

all would lose business, and all eight did in fact impose the conditions.140

In this variant of tacit agreement, in other words, the rivals interdependently accept a common

proposal by a trading partner, thereby reducing competition to their joint benefit.141 The decisive

139 Id. The plaintiff in Cities Service itself, by contrast, failed to show that the defendant oil company

had any incentive to participate in an alleged boycott, so the same inference from parallel conduct was not

appropriate. See also United States v. Citizens & S. Nat’l Bank, 422 U.S. 86, 113 (1975) (describing the

plaintiff’s argument that a bank’s “correspondent associate programs have actually encompassed at least a

tacit agreement of fix interest rates and service charges” and citing Interstate Circuit, Masonite, and Bausch

& Lomb).

140 White v. R.M. Packer Co., 635 F.3d 571, 576 (1st Cir. 2011).

141 See, e.g., Toys “R” Us, Inc., v. FTC, 221 F.3d 928, 936 (7th Cir. 2000) (finding sufficient

evidence to infer conspiracy of toy manufacturers arranged by Toys “R” Us (TRU); the FTC “was not

required to disbelieve the testimony of the different toy company executives and TRU itself to the effect

that the only condition on which each toy manufacturer would agree to TRU’s demands was if it could be

sure its competitors were doing the same thing”); Meyer v. Kalanick, 15 Civ. 9796, 2016 WL 1266801, at

*4 (S.D.N.Y. Mar. 13, 2016) (finding plaintiff sufficiently a hub-and-spokes agreement in which “drivers

agree with Uber to charge certain fares with the clear understanding that all other Uber drivers are agreeing

to charge the same fares”); In re Nexium (Esomeprazole) Antitrust Litig., 42 F. Supp. 3d 231, 254 (D.

Mass. 2014) (“Like the agreements at issue in Interstate Circuit . . . AstraZeneca’s agreements with the

Generic Defendants demonstrate a degree of interdependence suggesting a single agreement, even if no

such agreement was expressly made between the Generic Defendants.”); Lauman v. NHL, 907 F. Supp. 2d

465, 486–87 (S.D.N.Y. 2012) (“Plaintiffs do not plausibly allege that the [regional sports networks, or

RSNs] entered into actual agreements with one another to enforce the territorial market divisions

established by the League defendants, but it is not necessary that they do so in order to implicate the RSNs

in the conspiracy to divide the market.”); Columbus Drywall & Insulation, Inc. v. Masco Corp., No. 04–

CV–3066, 2009 WL 856306, at *12–13 (N.D. Ga. Feb. 9, 2009) (finding sufficient evidence to support a

claim that an insulation contractor “orchestrated and facilitated an agreement between [insulation]

40

issue in cases like this is whether the rivals act interdependently because of the vertical agreements;

it is insufficient that they act in the same way, if they had independent reasons for doing so.142

Interstate Circuit is usually cited as an instance of a hub-and-spokes conspiracy: the

arrangement involves both express vertical agreements to form the spokes and a tacit horizontal

agreement to form the rim. In cases like this, a single party, usually (although not always)143 one

manufacturers to maintain and increase [the contractor’s price advantage over its rivals] in exchange for

[the contractor’s] agreement to support industry-wide price increases”; the contractor “communicat[ed]

directly with the manufacturers, and sometimes demand[ed] price adjustments, when it received reports”

of violations, sometimes from the manufacturers themselves).

Similarly, in United States v. Apple, Inc., 791 F.3d 290, 316 (2d Cir. 2015), cert. denied, 136 S. Ct.

1376 (2016), the court found sufficient evidence that “Apple consciously orchestrated a conspiracy among”

ebook publishers to increase the retail price of ebooks by interdependently switching to an agency model

of distribution with a most-favored-nation clause, requiring publishers to set a retail price on Apple’s

bookstore no higher than its price to other retailers. In this case, however, the court found that Apple went

beyond the Interstate Circuit scenario by “consciously play[ing] a key role in organizing [the publisher’s]

express collusion,” id. at 318, to the point of “coordinat[ing] phone calls between the publishers who had

agreed and those who remained on the fence.” Id. at 319.

142 In re Ins. Brokerage Antitrust Litig., 618 F.3d 300, 331 (3d Cir. 2010) (rejecting plaintiffs’

reliance on Interstate Circuit and a hub-and-spokes theory because the defendants had not acted

interdependently in response to the common proposal); Barry v. Blue Cross of Cal., 805 F.2d 866, 869 (9th

Cir. 1986) (finding no “tacit conspiracy” under Interstate Circuit, because that there was no showing that

the [defendants] were economically interdependent—that ‘cooperation was essential to successful

operation of the plan’”) (internal citation omitted); Radio Music License Comm., Inc. v. SESAC, Inc., 29

F. Supp. 3d 487, 497–98 (E.D. Pa. 2014) (holding that pattern of licensing of public performance rights to

radio stations did not establish a hub-and-spokes conspiracy among the stations, because each station had

an independent reason accept the licensing terms). In In re Musical Instruments & Equipment Antitrust

Litigation, 798 F.3d 1186 (9th Cir. 2015), however, the court dismissed insufficient allegations of a hub-

and-spokes conspiracy, even though the defendants responded interdependently to a powerful retailer’s

demand for adoption of “minimum advertised price” policies on all retailers, reasoning that

“[m]anufacturers’ decisions to heed similar demands made by a common, important customer do not

suggest conspiracy or collusion. They support a different conclusion: self-interested independent parallel

conduct in an interdependent market.” Id. at 1195. In essence, the majority in Musical Instruments treated

the common proposals from a common downstream firm no differently than any other market event, so an

interdependent response was not concerted. It did not cite Interstate Circuit. Even at that, a dissent found

sufficient allegations to make an agreement plausible. Id. at 1198–99 (Preferson, J., dissenting) (“When

truly analyzed together, the six plus factors strongly suggest that the manufacturer defendants reached an

illegal horizontal agreement, which ‘nudge’ plaintiffs' allegations ‘from conceivable to plausible.’”).

143 In re Nexium (Esomeprazole) Antitrust Litig., 42 F. Supp. 3d 231, 253 (D. Mass. 2014), held

that Interstate Circuit’s logic applied to a series of proposed pay for delay settlements between a branded

drug manufacturer and its generic rivals. Interstate Circuit required only that “the parties acquiescing to the

proposed arrangement . . . be direct competitors.” Id. One of the rivals was found to have acquiesced by

“sign[ing] an agreement to stay out of the generic Nexium market as long as its competitors did the same.”

41

in a vertical relationship to the rivals, makes a simultaneous proposal to all rivals, which the rivals

accept explicitly, uniformly, and interdependently, to their joint benefit.144 Nevertheless, it makes

sense to treat at least some hub-and-spokes conspiracies as a species of tacit agreement, because

express agreements may serve a function similar to a direct communication of intention among

rivals; the agreements coordinate the interdependent responses of the rivals by providing strategic

information that each recipient knows its rivals are also receiving. The pattern of responses differs

from simple interdependence because the form of the communication to each of the upstream

suppliers, identifying all of the recipients, serves a function similar to a horizontal communication

of the same proposal.

III. PUBLIC COMMUNICATIONS ABOUT PRESENT ACTIONS

Public announcements of present prices, even if they coordinate noncompetitive pricing,

are insufficient to satisfy the conditions for a tacit agreement.145 The Supreme Court pointed to

one reason in Brooke Group, when it noted that coordinating a price increase “through the

conscious parallelism of oligopoly must rely on uncertain and ambiguous signals to achieve

concerted action,” a “minuet” that is “difficult to compose and to perform, even for a disciplined

Id. at 259. That was “compelling evidence of acquiescence to a common scheme to delay generic

competition” even though its “acquiescence may not have had any effect on the state of generic

competition.” Id.

144 If the agreement is purely for the benefit of the party making the proposal, then there is little

basis for inferring an agreement. Turner, supra note 11, at 701 (“It would seem somewhat strange to call

this a horizontal agreement [because] each ended up worse off than before. An agreement among

competitors on a course of conduct with such unhappy consequences would be an odd sort of agreement.”).

Such a scenario would be better evaluated under Section 2 of the Sherman Act. See id. at 700–03 (discussing

Lorain Journal Co. v. United States, 342 U.S. 143 (1951)).

145 See, e.g., Rigby v. Philip Morris USA, Inc., No. CV 513–110, 2015 WL 1275412, at *6 (S.D.

Ga. Mar. 19, 2015) (dismissing allegations as insufficient to infer collusive conduct where the defendants

published a price sheet in early 2009 and then later re-issued a price sheet with lower prices that were “in

line” with those offered by defendants’ competitor).

42

oligopoly.146 In a one-shot game, a price increase is unlikely to convey enough information about

the sender’s conciliatory intent for rivals to coordinate such a price increase. But oligopolists in

real-world markets do not play one-shot games. In repeated games, it is possible that rivals can

indicate their intention by patterns of price movements and responses and eventually achieve tacit

collusion.147 The pattern of price announcements by gas stations on Martha’s Vineyard, described

in White, is one real-world example among many: the stations admitted that they communicated

indirectly by posting their prices and were able to engage in cooperative pricing by “rationally”

predicting what rivals would do in response.148

146 Brooke Grp. Ltd. v. Brown & Williamson Tobacco Corp., 509 U.S. 209, 227–28 (1993). Justice

Stevens objected that “professional performers who had danced the minuet for 40 to 50 years would be

better able to predict whether their favorite partners would follow them in the future than would an outsider,

who might not know the difference between Haydn and Mozart.” Id. at 257.

147 Experiments suggest that if there are very few rivals, many buyers, visible pricing, and

homogeneous products, and firms compete over many time periods, each firm may be able to predict

whether its rivals will likely reciprocate a price increase and coordinate a collusive price. See, e.g., Damien

J. Neven, “Collusion” Under Article 81 and the Merger Regulation, KONKURRENSVERKET, SWEDISH

COMPETITION AUTHORITY, FIGHTING CARTELS–WHY AND HOW? 56, 65 (2001) (summarizing

experimental results suggesting that, without direct communications, successful collusion only appears

frequently “in very simple market environments . . . with only two firms and a lot of experience,” but if

“the number of players is increased, or the environment is made more complicated (for instance with

different costs or different demands across firms), firms appear to be unable to sustain collusive outcomes”).

This result is consistent with the Folk Theorem, which states that “if firms can observe each other’s actions

and interact with one another sufficiently frequently, then the firms can maximize their joint payoffs . . .

without communication or transfers.” MARSHALL & MARX, supra note 6, at 9. See also Green et al., supra

note 11, at 469 (“[A] standard folk theorem characterizes the set of equilibria of a repeated oligopoly game

and shows that for sufficiently patient firms (or for sufficiently short delay between repetitions of the game),

the set of equilibria includes strategy profiles that generate the monopoly outcome.”).

148 Brief for Defendants–Appellees, White v. R.M. Packer Co., 635 F.3d 571 (1st Cir. 2012) (No.

10-1130), 2010 WL 3213231, at *31–32. An econometric study of Italian gasoline pricing also found a

pattern of price announcements for gasoline can facilitate “sticky,” less competitive pricing. Patrick

Andreoli-Versbach & Jens-Uwe Franck, Endogenous Price Commitment, Sticky and Leadership Pricing:

Evidence from the Italian Petrol Market, 40 INT’L J. INDUS. ORG. 32, 35 (2015) (when the largest Italian

gasoline refiner announced a policy of making fewer and larger price changes, regardless of cost, others

eventually followed suit and market prices increased; an antitrust investigation ended with no evidence of

communication).

43

Even if rivals succeed in coordinating prices solely by public announcements of their

present prices, however, the result is not a tacit agreement. Courts view the public announcement

of prices by itself as a necesssary part of oligopoly pricing,149 and within its safe harbor, because

it provides the information sellers and buyers need to make essential choices.150 To hold otherwise,

they reason, would limit necessary information in the market and require courts to engage in an

inquiry approximating price regulation in order to determine liability.151 Economists (like the

Supreme Court in the passage I quoted earlier from Brooke Group) sometimes describe these sorts

of announcements of prices as signaling.152 Calling them signaling, however, does not change the

legal characterization of the conduct. As one court put it, “public announcement of a pricing

decision cannot be twisted into an invitation or signal to conspire; it is instead an economic reality

to which all other competitors must react.”153 A sufficiently complex pattern of announcements

149 See also Blomkest Fertilizer, Inc. v. Potash Corp. of Sask., 203 F.3d 1028, 1036 (8th Cir. 2000)

(“[E]vidence that the alleged conspirators were aware of each other’s prices, before announcing their own

prices, is nothing more than a restatement of conscious parallelism, which is not enough to show an antitrust

conspiracy.”) (internal quotation marks omitted); Joseph Kattan, Beyond Facilitating Practices Price

Signaling and Price Protection Clauses in the New Antitrust Environment, 63 ANTITRUST L.J. 133, 139

(1994) (observing that “[f]irms must be able to react to market conditions, including the prices charged by

rivals”; a firm cannot “put blinders on and pretend that it does not know what its rivals are charging”;

consequently “the interdependent outcome may be unavoidable”).

150 See Kattan, supra note 149, at 138–39 (explaining that a simple price increase, without more,

cannot “be viewed predominantly as a signal to competitors rather than as a notice to customers”).

151 Posner, Kaplow Review, supra note 11, at 762–63.

152 See, e.g., George J. Mailath, Simultaneous Signaling in an Oligopoly Model, 104 Q.J. ECON.

417, 417 (1989) (presenting “a dynamic model of differentiated oligopoly in which all firms have private

information about their costs and simultaneously signal their information by their pricing decisions”)

(emphasis in original); Austin C. Hoggatt et al., Price Signaling in Experimental Oligopoly, 66 AM. ECON.

REV. 261, 262 (1976) (“Signals are special [i.e., unusual or noteworthy] price choices which are chosen to

substitute for verbal communication when the latter is impossible.”).

153 United States v. Gen. Motors Corp., No. 38219, 1974 WL 926, at *21 (E.D. Mich. Sept. 26,

1974). See also Williamson Oil Co., Inc. v. Philip Morris USA, 346 F.3d 1287, 1310 (11th Cir. 2003)

(“None of the [price changes] that appellants label ‘signals’ tend to exclude the possibility that the primary

players in the tobacco industry were engaged in rational, lawful, parallel pricing behavior that is typical of

an oligopoly.”); Valspar Corp. v. E.I. Du Pont De Nemours, No. 14–527–RGA, 2016 WL 304404, at *10

(D. Del. Jan. 25, 2016) (“[P]arallel price increases, or ‘signals,’ would perhaps describe how a conspiracy

44

and price increases may support an inference that rivals coordinated the price increases by private

communication, particularly if there is evidence of meetings,154 but a series of identical announced

price increases is normally explicable as simple interdependence and insufficient to raise the

necessary inference of agreement.155

Courts might find agreement based on public announcements of present prices if the

particular announcement includes information that conveys a different, unique meaning useful to

rivals. Bidders in open-price, ascending spectrum auctions of regional licenses, for example, have

signaled their collusive intent to rivals using codes in bids.156 The bids are present offers to buy

practically functioned, but only if there were there some indication of an agreement to begin with, rather

than conduct that could just as well be explained by independent action. In short, nothing about these

announcements tends to exclude the possibility of independent action.”); Hall v. United Air Lines, Inc., 296

F. Supp. 2d 652, 670 & n.23 (E.D.N.C. 2003) (evidence that airlines cut base commissions to travel agents

“assum[ing] the contemplated cut would not be sustainable without competitor matching” and “hop[ing]

that others would match the cut” did not “mean that the defendants agreed to cut commissions by ‘offer and

acceptance’ as plaintiffs contend”) (citing General Motors). But cf. In re Titanium Dioxide Antitrust Litig.,

959 F. Supp. 2d 799, 828 (D. Md. 2013) (“Frequent price increase announcements could have served as

‘signals,’ making further exchange of actual price information superfluous.”); In re Currency Conversion

Fee Antitrust Litig., 773 F. Supp. 2d 351, 356–58, 371 (S.D.N.Y. 2011) (denying summary judgment where

“price announcements may have served as price beacons to competitors for the purpose of gauging their

willingness to raise prices” but there [also] was evidence of a relevant meeting among competitors’ in-

house counsel).

154 See Titanium Dioxide, 959 F. Supp. at 825 (“The sheer number of parallel price increases, when

coupled with the other evidence in this case, could lead a jury to reasonably infer a conspiracy.”); In re

Ethylene Propylene Diene Monomer (EPDM) Antitrust Litig., 681 F. Supp. 2d 141, 169, 176 (D. Conn.

2009) (denying summary judgment where plaintiffs offered evidence of “six ‘lockstep,’ industry-wide price

increases” and “frequent and friendly” communications and secret meetings among the defendants”).

155 See, e.g., In re Chocolate Confectionary Antitrust Litig., 801 F.3d 383, 400–01 (3d Cir. 2015)

(“[E]vidence of a price increase disconnected from changes in costs or demand only raises the question:

was the anticompetitive price increase the result of lawful, rational interdependence or of an unlawful price-

fixing conspiracy?”); In re Text Messaging Antitrust Litig., 782 F.3d 867, 871 (7th Cir. 2015) (“It is true

that if a small number of competitors dominates a market, they will find it safer and easier to fix prices than

if there are many competitors of more or less equal size. . . . But the other side of this coin is that the fewer

the firms, the easier it is for them to engage in ‘follow the leader’ pricing . . . which means coordinating

their pricing without an actual agreement to do so.”).

156 See United States v. Mercury PCS II, L.L.C., No. 98–2751 (PLF), 1999 WL 1425379, at *6

(D.D.C. April 29, 1999):

A bidder’s purpose in making a bid might, depending on the circumstances, be

ambiguous to its rivals. Where ambiguity remains, it can be difficult to use a bid or bidding

45

the available blocks of spectrum at the stated price per unit, but the bidders in a few instances

“use[d] the early rounds when prices are still low to signal their views about who should win which

objects, and then, when consensus ha[d] been reached, tacitly agree[d] to stop pushing prices

up.”157 In one instance, rivals achieved this goal by submitting bids with “trailing digits” in the

lower places of bid that corresponded to the regional market numbers of other auctions. The

otherwise arbitrary numbers would allegedly signal to rivals that they would face more aggressive

bidding for the designated markets if they persisted in bidding in the current auction. In effect, bids

signaled offers to refrain from bidding in the other auctions in return for the rivals’ forbearance in

pattern alone to send clear messages or invitations to collude. To eliminate or reduce any

ambiguity, Mercury sometimes placed bids during the DEF auction in which the final three

digits intentionally corresponded to the number for a BTA (a “BTA end code”). Knowing

that other bidders could see the bids and hence the BTA end codes, Mercury used the codes

to better explain the real purpose of certain bids it made—to reach an agreement with a

rival. In particular, Mercury used the BTA end codes to link the bidding of licenses in two

(or more) specific BTA markets, highlight the licenses Mercury wanted, and convey to the

competing bidders offers to agree with Mercury not to bid against each other for the linked

licenses.

The consent decree prevented “Mercury from using BTA end codes or any similar signaling mechanism to

solicit anticompetitive agreements in future FCC auctions.” Id. at *8; see also Peter Cramton & Jesse

Schwartz, Collusive Bidding: Lessons from the FCC Spectrum Auctions, 17 J. REG. ECON. 229 (2000). For

an analysis of bid rigging generally, see Robert C. Marshall & Michael Meurer, Bidder Collusion and

Antitrust Law: Refining the Analysis of Price Fixing to Account for the Special Features of Auction Markets,

72 ANTITRUST L.J. 83 (2004).

See also Dahl v. Bain Capital Partners, LLC, 937 F. Supp. 2d 119, 130–33, 142–45 (D. Mass. 2013)

(denying summary judgment based on what the court took to be a plausible fit with the White court’s

account of tacit agreement). Private equity firms had uniformly withdrawn bids in a leveraged buyout when

another firm announced a deal to buy the company. The court characterized the evidence as suggesting,

among other things, that the announcement of the deal was “request to the industry,” with which the rivals

complied by abruptly “standing down.” Id. at 142.

157 See PAUL KLEMPERER, AUCTIONS: THEORY AND PRACTICE 86 (2004). Cf. William G. Christie

& Paul H. Schultz, Why Do Nasdaq Market Makers Avoid Odd-Eighth Quotes?, 49 J. FIN. 1813, 1838–39

(1994) (identifying a suspiciously consistent pattern of market makers’ spreads between bid and ask prices

on the Nasdaq exchange). This article led to multidistrict antitrust litigation (eventually settled for over $1

billion) alleging that the spreads were the result of an explicit conspiracy. In re NASDAQ Market-Makers

Antitrust Litig., 187 F.R.D. 465, 492 (S.D.N.Y. 1998) (approving the settlement). A government

investigation ended in a consent decree approved by the same court. United States v. Alex. Brown & Sons,

Inc., 963 F. Supp. 235, 247 (S.D.N.Y. 1997), aff'd sub nom. United States v. Bleznak, 153 F.3d 16 (2d Cir.

1998).

46

the primary auction.158 In practical effect, then the purportedly public, present bid contained

private information about future actions, which took the bid out of the benign category.

In the foregoing auction, it was obvious to observers what the bidders were doing, so

regulatory responses—including antitrust challenges—were predictable. (In order to thwart this

strategy the FCC now constrains bidding increments and conceals the names of bidders during

auctions.)159 If an oligopolist were to announce a price change on its web site along with

comparable, easily decrypted code unmistakably suggesting a quid pro quo with rivals, presumably

the court would treat it in the same way as a private communication, despite visibility to others.

But that very visibility makes it far less effective in coordinating an illegal conspiracy than a truly

private communication.

IV. PRIVATE COMMUNICATIONS ABOUT PRESENT ACTIONS

In United States v. United States Gypsum Co. 160 the Supreme Court recognized that private

exchanges of price information may “increase economic efficiency and render markets more,

rather than less, competitive,” but cautioned that “[e]xchanges of current price information . . .

have the greatest potential for generating anticompetitive effects.”161 Courts distinguish whether

158 Werden, supra note 30, at 766–67 (observing that by this sort of auction mechanism “the

defendants communicated and reached an agreement on price . . . without an explicit offer or acceptance”;

and that “the communication was exceptionally minimal but nevertheless effective” in the same way as a

“traditional conspirac[y]”). An even clearer example occurred in a German auction of ten blocks of

spectrum. By rule, each new bid had to be at least 10% higher than the previous bid. Firm A initially bid

DM 20 million/MHz on blocks 1-5 and DM 18.18 million on blocks 6-10; Firm B then bid DM 20 million

on blocks 6-10 and did not bid on blocks 1-5; there were no further bids. It seems evident that A’s oddly

specific bid of 18.18 on blocks 6-10 was a proposal that, if B did not contest blocks 1-5, B could, under the

10% minimum rule, win blocks 6-10 at a bid at or over 18.18 x 1.1 = 19.998—a proposal that B happily

accepted by bidding DM 20 million on blocks 6-10 and nothing on blocks 1-5. KLEMPERER, supra note

157, at 104–05.

159 MARSHALL & MARX, supra note 6, at 201–02.

160 438 U.S. 422 (1978).

161 Id. at 441 n.16.

47

evidence the latter sort of exchange justifies the inference of (1) an agreement to fix or stabilize

prices, which is per se unlawful or (2) an agreement to exchange information, which is unlawful

under the rule of reason only if actually reduces competition.162 I consider here when either sort of

agreement might take the form of tacit agreement.

A. PRICE FIXING (OR OTHER PER SE ILLEGAL) AGREEMENTS

Private communications of present pricing information among rivals are less probative of

agreement on prices themselves than are communications of future price information, even if they

provide no direct benefit to consumers.163 In markets in which transaction prices are not public,

private exchanges of prices can provide sellers with information that serves a legitimate purpose

by thwarting deception by buyers. In Blomkest, for example, the court refused to find that evidence

of large and parallel price increases together with “nearly simultaneous price verifications” among

rivals raised an inference of agreement, because the “communications only concerned charges on

particular completed sales, not future market prices” and there was “no evidence to support the

inference that the verifications had an impact on price increases.”164

162 In Gypsum, the Court held that evidence that the parties have agreed to exchange prices is

insufficient to sustain a criminal conviction for price fixing, even if the exchange affected competition; the

government must also offer independent evidence that the rivals acted with knowledge of the likely

consequences. Id. at 441–44.

163 See, e.g., Omnicare, Inc. v. UnitedHealth Grp., Inc., 629 F.3d 697, 720 (7th Cir. 2011)

(“[C]irculation of generalized and averaged high-level pricing data, policed by outside counsel, [was] more

consistent with independent than collusive action.”). But see Rosefielde v. Falcon Jet Corp., 701 F. Supp.

1053, 1064 (D.N.J. 1988) (finding that exchanges of current price information “facilitated a tacit agreement

among business jet manufacturers to adhere to the exchanged prices,” where there was some evidence of

that “senior executives . . . were aware of the price information exchange and considered the data obtained

by the sales engineers to set the price of business jets”).

164 Blomkest Fertilizer, Inc. v. Potash Corp. of Sask., 203 F.3d 1028, 1034 (8th Cir. 2000). See also

In re Baby Food Antitrust Litig., 166 F.3d 112, 125–26 (3rd Cir. 1999) (rivals’ collection of one another’s

present pricing information from a variety of sources did not raise an inference of agreement).

48

In another recent case, hospitals repeatedly shared current (and in some instances

“unmasked, competitively sensitive”) information about nurses’ compensation165 and only

“occasionally” exchanged information about “projected future wage increases.”166 The court

found, in a careful opinion that identified the competing considerations, that the evidence was

barely insufficient to raise an inference of per se illegal price fixing.167 The hospitals undoubtedly

wanted to know rivals’ wages in setting their own, but the court found insufficient evidence that

they were using that information to set wages cooperatively. The hospitals had a legitimate need

for comparative information in order independently to meet their “target” pay rates—individually

determined percentages of the highest wage in the market. Moreover, their wage practices were

not uniform, and their internal records demonstrated that each hospital used the shared information

to make “a series of discrete, individualized RN compensation decisions.”168 The court concluded

that “the evidence here simply features too wide a disparity among the Defendant hospitals’

processes for determining RN compensation and the outcomes of these processes to support a

reasonable inference of a conspiracy in violation of § 1.”169

B. INFORMATION EXCHANGE AGREEMENTS

In Foley, the court found that a one-time private exchange of pricing intentions, combined

with evidence that the rivals later acted consistently with their stated intentions, in some instances

165 Cason-Merenda v. Detroit Med. Ctr., 862 F. Supp. 2d 603, 641 (E.D. Mich. 2012). Rivals

engaged in “fairly regular surveys augmented by direct contacts.” Id. at 633. Although low-level employees

made the contacts, those with pricing authority used the information. Id.

166 Id. at 631.

167 Id. at 639.

168 Id. at 640.

169 Id. at 641. The court found, however, that the evidence was sufficient to raise a triable issue that

the hospitals had formed an agreement to exchange wage information that was illegal under the rule of

reason. Id. at 643. That part of the opinion is discussed in the next Part of this article.

49

after prodding, was sufficient to raise an inference of price-fixing.170 A one-time exchange of

present pricing information is far less probative of an agreement of any kind. If the exchanges are

frequent, however a court may infer an agreement to exchange information. In cases like these, the

court treats the agreement to exchange information as the relevant one and evaluates it under some

version of the rule of reason based on its actual effect on competition.171 The case then turns on

whether the information allows rivals to restrict output rather than to enable efficient allocation of

resources.172

In United States v. Container Corp.,173 for example, rivals periodically asked one another

for their most recent prices, and each rival “on receiving that request usually furnished the data

with the expectation that it would be furnished reciprocal information when it wanted it.”174 The

Court unanimously held that the “concerted action” in exchanging prices was “sufficient to

establish the combination or conspiracy, the initial ingredient of a violation of § 1 of the Sherman

Act.”175 The majority held it unlawful because the effect of the agreement was to stabilize prices

“at a downward level,”176 although a higher one than would have prevailed without the exchanges.

170 United States v. Foley, 598 F.2d 1323, 1332–33 (4th Cir. 1979).

171 Todd v. Exxon Corp., 275 F.3d 191, 198 (2d Cir. 2001) (holding that, in cases where “the

violation lies in the information exchange itself—as opposed to merely using the information exchange as

evidence upon which to infer a price-fixing agreement [—the agreement to exchange information] is not

illegal per se, but can be found unlawful under a rule of reason analysis”).

172 United States v. Citizens & S. Nat’l Bank, 422 U.S. 86, 113 (1975) (holding that “the

dissemination of price information is not itself a per se violation”); see also United States v. United States

Gypsum Co., 438 U.S. 422, 441 n.16 (1978) (describing the “the structure of the industry involved and the

nature of the information exchanged” as part of the analysis under the rule of reason).

173 393 U.S. 333 (1969).

174 Id. at 335.

175 Id.

176 Id. at 336–37; id. at 337 (“Stabilizing prices as well as raising them is within the ban of [§] 1 of

the Sherman Act.”).

50

A concurring opinion referred to the arrangement as a tacit agreement,177 apparently because the

repeated compliance with requests for information strongly suggested an intent to reciprocate; it

reduced price competition, because it “made it possible for individual defendants confidently to

name a price equal to that which their competitors were asking.”178 The dissent, however, thought

“the evidence establishe[d] that the information was used by defendants as each pleased and was

actually employed for the purpose of engaging in active price competition”179 and had no

“significant anticompetitive effect.”180

More recently, as I showed in the last Part, Cason-Merenda found that hospitals’ extensive,

“on-demand” sharing of current wage information did not justify an inference of an agreement to

fix prices, because other evidence suggested that the hospitals used the information to make

independent pricing decisions. The court did find, however, that the evidence was sufficient to

infer “that the Defendant hospitals conspired among themselves to exchange wage information in

a manner that harmed competition by depressing RN wages.”181 The court pointed to evidence of

exchanges of detailed, disaggregated information; expert testimony that these sorts of exchanges

can facilitate price coordination; expert testimony that wages were in fact below competitive

levels; and evidence that hospital executives “expressly referenced market surveys of RN wages

in explaining decisions to award smaller-than-recommended pay increases.”182 Such an agreement

177 Id. at 340. See also In re Coordinated Pretrial Proceedings in Petroleum Prods. Antitrust Litig.,

906 F.2d 432, 447 (9th Cir. 1990) (citing Container as an instance of tacit agreement inferred from price

exchanges).

178 Container, 393 U.S. at 339–40 (Fortas, J. concurring).

179 Id. at 344 (Marshall, J., dissenting).

180 Id. at 346–47.

181 Cason-Merenda v. Detroit Med. Center, 862 F. Supp. 2d 603, 641 (E.D. Mich. 2012).

182 Id. at 645. The court also noted that the exchanges did not comply with the “safe zone” described

in the enforcement agencies’ statement of policy on “provider participation in exchanges of price and cost

information.” See U.S. Dep’t of Justice & Fed. Trade Comm’n, Statements of Health Care Antitrust

51

could be anticompetitive if it allowed each hospital to “hold the various elements of its RN

compensation package to the minimum level needed to meet its objectives, secure in the

knowledge that it did not need to outbid local competitors whose wage rates and practices were

unknown.”183 Although the tacit agreement was only to exchange the information and not to limit

price competition, there was sufficient evidence reasonably to infer that the agreement in practice

had anticompetitive effects.

V. PUBLIC COMMUNICATIONS ABOUT FUTURE ACTIONS

In his discussion of “prior announcements of moves,” Michael Porter observes that “an

entire competitive battle can be waged through announcements before a single dollar of resources

is expended,”184 citing an example in which a firm persuaded its rivals not to produce a competing

product by making a series of announcements of prices “to be available two years hence.”185 In

the same way, “trading announcements back and forth can settle the size of a price change . . .

without the need to disrupt the market and risk a battle by actually introducing one scheme and

then having to change or withdraw it later.”186 This process may sound a lot like the realtors’ serial

announcements of their intentions in Foley, with a similar outcome. Nevertheless, courts have

rejected that characterization, either because the announcements are too ambiguous to serve the

Enforcement Policy, Statement 6 (revised Aug. 1996), www.justice.gov/atr/statements-antitrust-

enforcement-policy-health-care#CONTNUM_49. The safe zone requires that the information be

aggregated from at least five providers , gathered by a third party, and at least three months old. Cason-

Merenda, 862 F. Supp. 2d at 612–15.

183 Id. at 645.

184 PORTER, supra note 41, at 76–78.

185 Id. at 78–79.

186 Id. at 79.

52

purported signaling function,187 or because the announcements serve benign purposes and may

benefit multiple audiences, especially consumers.188 In the DuPont (Ethyl) case,189 for example,

the court recognized that advance price announcements made it easier for rivals to coordinate their

pricing decisions, but also “aid[ed] buyers in their financial and purchase planning.”190

In rare instances, however, the content and context of a formal public announcement may

make it functionally private, without a comparable consumer benefit. First, if buyers have no use

for the price information, a court may find that rivals are the only audience. If rivals announce

prices in advance when their buyers are under long-term contracts, for example, the only apparent

purpose and effect is to negotiate a price change.191 Rivals may be able to do the same thing by

187 See also Reserve Supply Corp. v. Owens-Corning Fiberglas Corp., 97 F.2d 37, 54 (7th Cir.

1992) (rivals’ advance price announcements “reduce[d] the uncertainty inherent in raising prices by

allowing competitors time to decide if they would follow suit” and thus “facilitated the parallel pricing that

occurred in the market,” but did not permit an inference of price fixing, because they also provided

customers with the information they needed to “bid on building contracts well in advance of starting

construction”).

188 See, e.g., Blomkest Fertilizer, Inc. v. Potash Corp. of Sask., 203 F.3d 1028, 1037 (8th Cir. 2000)

(characterizing these actions as simply dissemination of price information and “far too ambiguous to defeat

summary judgment”).

189 E.I. du Pont de Nemours & Co. v. FTC, 729 F.2d 128, 133–34, 140–42 (2d Cir. 1984) (finding

no violation of Section 5 of the Federal Trade Commission Act when defendants’ use of advance price

announcements, price protection clauses, and delivered pricing had independent justifications and benefited

consumers).

190 Id. at 134. See also Kenneth G. Elzinga, New Developments on the Cartel Front, 29 ANTITRUST

BULL. 3, 13 (1984) (“The stubborn fact remains that in product markets buyers commonly find price lists

to be useful data, and many buyers will testify that their planning is facilitated and business risk is reduced

when prices do not change without advance notice from sellers.”).

191 In In re Coordinated Pretrial Proceedings in Petroleum Products Antitrust Litigation, 906 F.2d

432, 448 (9th Cir. 1990), for example, refiners publicly announced new wholesale prices of gasoline,

sometimes in advance, even though the information was useless to dealers, who had long-term supply

contracts with individual refiners. According to the court, the refiners’ officers testified that they announced

the prices “for the purpose of quickly informing competitors of the price change, in the express hope that

these competitors would follow the move and restore their prices.” Id. at 446. Unlike announcements of

retail prices, these announcements did not “foster[] the efficient or rational operation of the market

inasmuch as there simply was no wholesale market to rationalize; the branded dealers were not free to shop

around for their oil.” Id. at 448. But see In re Citric Acid Litig., 191 F.3d 1090, 1096 (9th Cir. 1999)

(characterizing this reasoning as dicta because the result in the case was also supported by direct evidence

of conspiracy).

53

preannouncements in spot retail markets, but, in that case, there is an efficiency justification in the

benefits to consumers of being able to plan for upcoming pricing changes. Second, if the

announcement includes additional information and conditions that make it useful only to rivals,

then it might be interpreted solely as a private signal of intention. If, for example, airlines serially

announce proposed fares that they link by footnote designations to form a bundle, and if consumers

cannot even book travel at the proposed fares, then one might infer that the purpose of the

announcements was only to communicate with rivals to negotiate an agreement.192 If, by contrast,

the airlines were simply to announce real fare increases for future travel, the same inference that

the communications were functionally private would not be justified; a resulting parallel fare

increase would not reflect a tacit agreement.193

192 The Antitrust Division alleged that airlines coordinated fares through their jointly operated

computer reservation system; the airlines eventually terminated the practice under a consent decree. United

States v. Airline Tariff Publ’g Co., 836 F. Supp. 9, 13 (D.C. Cir. 1993) (approving consent decree and

agreeing with the government that the terms of the fare announcements were of virtually no use to

consumers). Indeed, the government argued that the information was sufficiently detailed to amount to

assurances and the pattern of announcements negotiated an express agreement. The Competitive Impact

Statement stated that the “negotiation process continued until all significant airlines were lined up with the

same proposed fare increase and the same first ticket date, thus providing each other with commitments and

assurances as to the amount, scope, and timing of the proposed fare increase.” United States v. Airline

Tariff Publ’g Co., Proposed Final Judgment and Competitive Impact Statement, 59 Fed. Reg. 15,225,

15,230–31 (Mar. 31, 1994). See also Baker, supra note 84, at 51–52 (1996) (“[T]he Antitrust Division

alleged that the course of conduct amounted to an illegal agreement—actually many illegal agreements

over various fares and routes—and included elements that could and should be enjoined.”). For further

discussion, see Werden, supra note 30, at 765–66, and Severin Borenstein, Rapid Price Communication

and Coordination: The Airline Tariff Publishing Case, in THE ANTITRUST REVOLUTION 233 (John E.

Kwoka, Jr. & Lawrence J. White eds., 4th ed. 2004).

193 The Antitrust Division explained that the decree did not prohibit announcing “bona fide fares”

then withdrawing them if other airlines did not match them, but maintaining them if they did, because doing

so was simply taking account of “publicly available information,” a term that “encompasses information

concerning other airlines’ current and prior bona fide fares and fare changes, as well as any ‘pattern’ that

emerges from changes in such fares.” Proposed Consent Decree and Competitive Impact Statement, 59 Fed.

Reg. at 15,235–36. See Kühn et al., supra note 93, at 185 (explaining that, under the Consent Decree, “the

quoted prices generally have real effects and become commitments towards consumers [and] the immediate

availability of booking data does have efficiency effects in facilitating capacity planning [, so] it is probably

not sensible to restrict price announcements by firms that have some degree of commitment effects towards

the consumer”).

54

The medium or setting of an announcement can also indicate that its audience is primarily

rivals. “An announcement in a specialized trade journal,” Porter notes, “is likely to be noticed only

by competitors or other industry participants [and so] may carry a different connotation from

announcement made to a broad audience of security analysis or to the national business press.” 194

The same could be said about speeches at trade association meetings,195 press releases,196 pricing

letters,197 statements in trade journals, and earnings calls or meetings with stock analysts.198 Rivals

can sometimes offer far more detailed explanatory information to price announcements or other

194 PORTER, supra note 41, at 80

195 See, e.g., In re Travel Agency Comm’n Antitrust Litig., 898 F. Supp. 685, 691 (D. Minn. 1995)

(finding that evidence of “a set of occurrences, speeches, meetings, events, official and unofficial corporate

utterances, and conferences at which information was exchanged” was sufficient to raise an inference that

one airline signaled that it would cut broker commissions if others followed and the others signaled that

they would follow); Fears v. Wilhelmina Model Agency, Inc., No. 02 Civ. 4911(HB), 2004 WL 594396, at

*8 (S.D.N.Y. Mar. 23, 2004) (finding that advance announcements of commission increases at trade

association meetings, in circumstances in which other communications were occurring, may be interpreted

less as taking “the lead in raising its commissions, with the blind hope that other members would follow”

than as seeking “some guarantee that the other defendants would acquiesce, to ensure that its clients would

not segregate against them”).

196 Travel Agency Commission, 898 F. Supp. at 690

197 Similar considerations apply to other formally public, detailed communications that primarily

address rivals. For example, “pricing letters,” ostensibly to customers, may offer rivals detailed plans for

anticipated increases in prices. McWane, Inc. & Star Pipe Prods., Ltd., Docket No. 9351, 2012 WL

4101793, at *12 (F.T.C. Sept. 14, 2012) (denying summary decision for defendants and noting that the

language in the pricing letters “would have been meaningless to customers”). The alleged conspiracy in

McWane included the creation of an information exchange to monitor rivals’ pricing. Id. at *16–17. The

Commission later split 2-2 on liability on these theories, and so dismissed the counts without opinion.

McWane, Inc. & Star Pipe Prods., Ltd., Docket No. 9351, 2014 WL 556261, at *1 (F.T.C. Jan. 14, 2014)

(“[T]wo Commissioners find that Counts 1 and 2 alleging an unlawful conspiracy and information exchange

have been proven and two Commissioners do not. In the absence of a majority decision, we dismiss these

counts in the public interest.”).

198 See, e.g., In re Plasma-Derivative Protein Therapies Antitrust Litig. 764 F. Supp. 2d 991, 1001

(N.D. Ill. 2011) (statements that a company is “in a very good position to see stable growth in this business

going forward over the next three to five years” and that “the industry now heading to a much more

predictable phase of stability” were “very general statements to investors or regulators about the trajectory

of the plasma therapies industry” that failed to make allegations of conspiracy plausible). Cf. Howard

Rosenblatt & Tomas Nilsson, Analyst Calls and Price Signaling under EU Law, ANTITRUST SOURCE, June

2012 (discussing enforcement actions under Section 5 of the Federal Trade Commission Act against firms

for anticompetitive announcements during telephone calls with analysts, but monitored by rivals), at 1–3,

www.americanbar.org/content/dam/aba/publishing/antitrust_source/jun12_full_source.authcheckdam.pdf

55

strategy choices in oral or written statements about competitive conditions and strategies. As Porter

explains:

It is not uncommon for competitors to comment on industry conditions, including

forecasts of demand and price, forecasts of future capacity, the significance of

external changes such as material cost increases, and so on. Such commentary is

laden with signals because it may expose the commenting firm’s assumptions about

the industry on which it is presumably building its own strategy. As such this

discussion can be a conscious or unconscious attempt to . . . minimize the chances

of mistaken motives and warfare. Such commentary can also contain implicit pleas

for price discipline: “Price competition is still very harsh. The industry is doing a

lousy job of passing along increased costs to the consumer.” “The problem in this

industry is that some firms do not recognize that these current prices will be

detrimental to our ability to grow and produce a quality product in the long run.”

Or discussions of the industry may contain implicit pleas that other firms add

capacity in an orderly fashion not engage in excessive advertising competition, not

break ranks in dealing with large customers, or any number of other things, as well

as implicit promises to cooperate if others act “properly.”199

Nevertheless, the statements provide legitimate information to a range of market constituencies,

so courts refuse to interpret them as signals in a relevant legal sense.200 These communications

199 PORTER, supra note 41, at 81 (internal citations omitted).

200 See, e.g., In re Nat’l Ass’n of Music Merchs., Musical Instruments & Equip. Antitrust Litig.,

MDL No 2121, 2012 WL 3637291, at *2 (S.D. Cal. Aug. 20, 2012) (“[U]nilateral advocacy, particularly in

an open and public forum, is not itself an agreement or conspiracy [and] independent responses to public

advocacy without an agreement, even if consciously parallel to other entities’ activity, would simply be

permissible parallel conduct.”), aff’d sub nom. In re Musical Instruments & Equip. Antitrust Litig., 798

F.3d 1186 (9th Cir. 2015). The district court added that “attendance at trade shows or other large meetings

56

provide critical information to the market that, in some instances, may be required by securities

laws.201

If they are sufficiently clear and include information of no use to consumers, however, and

result in interdependent action, they may arguably form tacit agreements.202 Because anyone,

including rivals, can usually listen in on earnings calls or read a transcript of them on the firm

website, they provide an opportunity to make far more detailed statements about competitive

strategy than a bare announcement of a future price increase. They allow a rival to discuss its

reasoning about future price and output decisions in a setting typically monitored by competitors

and generally not by consumers. Nevertheless, cases finding that rivals formed an agreement of

any kind by these means are rare. The district court opinion in Holiday Wholesale Grocery203

offered a typically benign view of communications in analyst calls. In a lengthy section on

purported signaling, the court considered the wholesalers’ allegations that cigarette manufacturers,

through an intermediary, monitored one another’s signals made during their presentations to

[does not] imply an agreement or conspiracy,” especially where they are “large meetings attended by

numerous other people, including press representatives” because “[t]he presence of numerous uninvolved

observers at such meetings tends to dispel any specter of illegality.” Id. Moreover, “making announcements

about new practices or developments is common and doesn’t imply illicit or surreptitious signaling was

going on.” Id. at *5. But see Jeffrey L. Kessler & Ronald C. Wheeler, An Old Theory Gets New Life: How

to Price Without Being a “Price Signaler”, 7 ANTITRUST 26, 28 (Summer 1993) (arguing that press releases

may be more suspicious than announcements “limited only to customers and others with a legitimate need

to know,” unless there are legitimate justifications for the use of the press release format).

201 For discussion, see Richard M. Steuer et al., The Application of Antitrust to Public Companies’

Disclosures, 63 ADMIN. L. REV. 159, 159–60 (2011) (arguing that all but the most plainly anticompetitive

statements should be immune from antitrust scrutiny).

202 Statements in an earnings call may also, according to the FTC, constitute an invitation to collude

in violation of Section 5. See, e.g., Consent Order, etc., in Regard to Alleged Violations of Sec. 5 of the

Federal Trade Commission Act, Valassis Commc’ns, Inc., Docket No. C-4160, 2006 WL 6679058, at *9,

141 F.T.C. 5 (F.T.C. Apr. 19, 2006) (“In its earnings call, Valassis communicated to rival News America

proposed terms of coordination for the FSI market, a longstanding duopoly, and did so with extraordinary

specificity.”).

203 Holiday Wholesale Grocery Co. v. Philip Morris, Inc., 231 F. Supp. 2d 1253, 1276 (2002), aff’d

sub nom. Williamson Oil Co. v. Philip Morris USA, 346 F.3d 1287 (11th Cir. 2003).

57

analysts to coordinate specific price increases.204 The court was at pains to distinguish the

connotation of the word signaling from the specific alleged conduct. The term insinuates,

according to the court, “either that a competitor is inviting all to make a traditional price-fixing

agreement,” apparently using something like what I am calling an implicit signal, or “that there

already exists such an agreement and it is being carried out through indirect communications”205—

what I am calling a prearranged signal.

Despite plaintiffs’ use of terms like nods and olive branches, their evidence, according to

the court, showed only that “in an oligopoly, each company is aware of the others’ actions” and

takes them into account in their competitive decisions.206 The court reasoned that “[b]ecause in

competitive markets, particularly oligopolies, companies will monitor each other’s

communications with the market in order to make their own strategic decisions, antitrust law

permits such discussions even when they relate to pricing.”207 For there to be an “inference of

traditional agreement from indirect communications,” there must be “communications with no

public purpose,”208 like “internal competitive memos” and “supply and demand forecasts.”209 Or,

in my terminology, the communications must be fully or functionally private and convey

information about future competitive choices.

204 Id. at 1275.

205 Id.

206 Id.

207 Id. at 1276.

208 Id. at 1277.

209 Id. at 1276–77. The court of appeals agreed with the district court’s benign interpretation of the

evidence. Williamson Oil Co. v. Philip Morris USA, 346 F.3d 1287, 1305–09 (11th Cir. 2003). The court

recognized that the firms’ public statements and actions conveyed implicit signals that may have influenced

their rivals’ pricing behavior and led to price increases; it disagreed that signaling by these means was

collusive.

58

In rare cases, however, statements in earnings calls may be specific enough and sufficiently

directed solely to rivals to raise an inference of agreement. In Delta/AirTran, for example, rival

airlines allegedly proposed in earnings calls and industry meetings to reduce output and raise fees

for travelers, then acted consistently with the proposals.210 One announced that “we strongly

believe that more industry capacity needs to be removed,”211 and its rival the next day stated it

wanted to remove capacity, but had “to do it in conjunction with the other carriers” and that “if the

industry could achieve a 10% reduction in capacity year-over-year by the fall that we’d be in pretty

good shape given today’s fuel environment.”212 They made similar statements concerning prices

and fees in subsequent earnings calls and industry meetings, then simultaneously increased fees

and reduced capacity during a recession in which oil prices were falling.213 The court concluded

that the complaint’s extraordinarily detailed account of the announcements and their effects on

competition sufficiently alleged a Section 1 agreement.214 The court distinguished the case from

210 In re Delta/AirTran Baggage Fee Antitrust Litig., 733 F. Supp. 2d 1348, 1352 (N.D. Ga. 2010).

211 Id.

212 Id. at 1353.

213 Id. at 1356.

214 Id. at 1363. Another court summarized the remarkable detail of the Delta/Airtran complaint:

For example, following AirTran’s initial invitation to Delta to collude, Delta

indicated during a public earnings call that it had no plans to implement a first-bag

fee. Thirteen days later, during AirTran’s public earnings call (which it knew Delta

monitored), AirTran stated that it was considering the viability of implementing

first-bag fees. Roughly two and a half months later, Delta indicated during another

public earnings call (which it knew AirTran monitored) that it was willing to

impose first-bag fees. Eight days later, AirTran remarked during its public earnings

call that it wanted to implement first-bag fees and had invested in the capability to

do so, but that it had not yet implemented the fee because Delta had not done so.

Less than two weeks later, Delta announced that it was implementing a $15 first-

bag fee, effective December 5, 2008. One week later, AirTran announced that it

was implementing the exact same fee on the exact same day.

Duty Free Ams., Inc. v. Estee Lauder Cos., 946 F. Supp. 2d 1321, 1332 n.8 (S.D. Fla. 2013) (citing

Delta/AirTran, 733 F. Supp. 2d at 1352–56). For a recent complaint identifying earnings calls as evidence

of collusion, see Complaint ¶¶ 68-76, International Union of Operating Engineers Local 30 Benefits Fund

59

allegations of “mere price announcements” on the ground that the complaint alleged “that each

Defendant signaled its willingness to cut capacity and increase prices if the other Defendant acted

in concert.”215 Although neither the parties nor the court used the term tacit agreement, the

sequence of announcements and actions in the case is better understood in those terms, because of

its dependence on both statement and implementation.216

VI. CONCLUSION

Courts and scholars have understood (and misunderstood) tacit agreement in different

ways, but it remains an important part of Section 1 doctrine. I propose here a clearer understanding

of the category, both in concept and proof. Starting with Twombly, I show the place of tacit

agreement in the hierarchy of means of coordination, distinguishing it especially from mere

interdependence on the one hand and express agreement on the other. As a starting point, I draw

from the case law a definition—interdependent conduct coordinated by prior private

“conversations” about competitive intentions—then suggested some clarifications of the forms of

v. Lannett Co., No. 16-cv-990, 2016 WL 821578 (E.D. Pa. Mar. 2, 2016) (alleging collusion in dramatic

increases in prices of generic drugs; citing, among other sources, statements during earnings calls in which

defendants predicted stable pricing and reduced competition for market share). But cf. Credit Bureau Servs.

v. Experian Info. Sols., Inc., No. 12-61360-CIV, 2012 WL 6102068, at *17–18 (S.D. Fla. Dec. 7, 2012)

(concluding that statements in earnings call and memoranda did not plausibly propose the alleged

agreement to put small resellers out of business; instead, they suggested only the legitimate need to promote

greater data security).

215 Delta/AirTran, 733 F. Supp. 2d at 1362. The FTC has found that these sorts of statements can

amount to an invitation to collude in violation of Section 5 of the Federal Trade Commission Act. See also

Fed. Trade Comm’n, U–Haul Int’l, Inc. and AMERCO, 75 Fed. Reg. 35,033, 35,034 & n.3 (June 21, 2010)

(in which U–Haul allegedly announced during an earnings call that it had raised rates and suggested that

Budget’s failure to match the price increases, within 3-5 percent, would be unfortunate and would require

U-Haul to cut its rates to maintain market share). Cf. Liu v. Amerco, 677 F.3d 489, 494 (1st Cir. 2012)

(holding the same conduct could support a private right of action for damages under a Massachusetts statute

that “parallels” the Federal Trade Commission Act).

216 See Delta/AirTran, 733 F. Supp. 2d at 1356 (quoting the complaint’s allegation that “instead of

refraining from implementing a first-bag fee (Delta had said in July that it would not plan to implement the

fee, notwithstanding the Northwest merger), Delta announced on November 5, 2008—less than two weeks

after AirTran’s statement that it would prefer to be a follower’ on the first-bag fee—that Delta would begin

charging passengers a $15 first-bag fee, effective December 5, 2008”).

60

communication implicit in that definition. For that purpose I pointed to cases that illustrate tacit

agreement under the definition, identifying both standard examples (like Foley) and hybrid

examples (like Interstate Circuit). Finally, I proposed a classification of communications among

rivals based on their probative value in the inference of tacit agreement, depending upon whether

the communications are public or private, and whether they relate to present or future conduct.

Both the clarified definition and the classification of categories of evidence of communications

will, I hope, assist litigants and courts in resolving issues of agreement at every stage of Section 1

litigation.